The news coverage surrounding the breach of the AI platform Hugging Face by OpenAI’s autonomous agents reads like a scene straight out of a Terminator movie (apparently, a prequel). Yet for enterprise security teams — even those that don’t develop or deploy artificial intelligence — the detailed incident report published by Hugging Face provides a critical operational roadmap. The incident needs to be evaluated on the basis of a single, fundamental question: “Which organizational security flaws, missing controls, configurations, or technologies allowed this attack to succeed?”

What happened?

During cyber-capability testing of OpenAI’s advanced AI models against the ExploitGym benchmark, an agent found a way to break out of its confinement, reached the internet, and breached the infrastructure of several companies — one of them being Hugging Face. The model reasoned that it could find solutions to the benchmark challenges on these external systems. Dozens of articles have since covered these incidents in detail, so no need to go over that here. What truly matters is what transpired within Hugging Face’s systems between July 9 and 13 while the AI agent operated unchecked inside. The Hugging Face incident report includes an interactive attack timeline, so we’ll just list the key stages. A malicious configuration in a dataset uploaded by the agent allowed it to exfiltrate credentials for a worker pod, and execute code within it. The agent then exploited cloud environment metadata and escaped the privileged pod to obtain root-level access on the host. It retrieved a substantial number of secrets from storage and used a stolen VPN key and shared cluster administrator credentials to pivot deeper into the internal network and source code repository before the Hugging Face security team blocked its access.

How significant is the fact that this was next-generation AI?

Regardless of the specific area of ​​activity of a company, tomorrow it may become a target of human attackers, fully autonomous systems, or AI agents manually controlled by human hackers.

Many AI practitioners have rushed to dismiss this incident based on assumptions that OpenAI will never allow its tech to get into serious hacking. However, security researchers have argued for months that even currently available open-source models can be used to launch both offensive research and real-world attacks. Human oversight, while still a requirement, can be minimal. Even mid-tier models can easily iterate through a dozen vulnerabilities, inventory a compromised infrastructure, and quickly figure out how to expand their reach across a network. Defenders must analyze incidents like this in detail, map the key findings to their own environments, and adapt security controls to counter both stealthy human actors and rapid, noisy AI agents.

Three distinguishing features of AI-powered attacks

The techniques detailed in the Hugging Face report are not novel — the agent didn’t invent anything new. However, three factors regarding the attack’s speed and scale fundamentally shift the economics for both attackers and defenders:

• Trial and error becomes significantly cheaper. Where a human hacker might test five attack vectors, an agent can iterate through five hundred. Scenarios that security teams previously dismissed as “theoretically possible, but impractical and unlikely” and placed them on the backlog can now become active threats.
• Attack execution and incident response windows shrink. High-speed attacks are not unheard of, while some ransomware operators achieve compromise within hours even without AI. However, this accelerated pace will likely become the new baseline in cybersecurity, as demonstrated by the agent obtaining admin privileges at Hugging Face within 13 hours.
• Numerous alerts and log entries generated by the agent’s trial-and-error attempts can both help and hinder attackers as well as defenders. This noise can serve both as an indicator for threat detection, and as camouflage to hide meaningful actions. It also acts as a natural bottleneck for incident response when the workflows are fully manual.

How to protect corporate infrastructure

Don’t ignore alerts from triggered defensive controls. This classic mistake — often highlighted by our incident response experts — was made at Hugging Face as well. Malware execution, port scanning, WAF triggers, requests to non-allowlisted resources, and failed authentication attempts are likely some of the event types that are blocked automatically. However, all these events are often categorized by SIEM systems as purely informational and routinely ignored. In the Hugging Face incident, numerous server-side request forgery (SSRF) attack attempts were successfully blocked over several days, yet no one batted an eyelash. Much like persistent human adversaries, the AI agent continuously retried until it succeeded. To detect such attacks, establish clear rules and triage workflows for defensive alert telemetry to ensure that priority automatically escalates when event volume spikes.

Investigate rapidly escalating error counts. Legitimate operations rarely generate hundreds of authentication failures within 10 minutes across four distinct systems. Analyze historical baselines, and generate medium or high-severity alerts depending on the deviation from normal error rates. As part of automated response protocols, you can temporarily block the user or system that is generating the anomalous activity, or enforce rate-limiting on them.

Configure escalation paths and conduct security team drills. Hugging Face’s correlation engine successfully detected the attack, yet it failed to grab the human operators’ attention. During security drills, organization must evaluate how quickly key personnel respond to critical incidents — even when these happen during off-hours such as at night on weekends.

Isolate services — internal ones included — that process untrusted input by design. The primary vulnerability at the start of the attack was a node configured to process machine learning datasets, which are widely known for potentially containing executable code. This node, capable of executing untrusted code, operated without a sandbox or adequate isolation controls. The risk extends beyond machine learning — to document conversion engines, image processing pipelines, user file upload handlers, and any other server-side processes that perform deserialization or template rendering. Any server-side execution of this nature must run under maximum feasible isolation, restricted from accessing external or internal network hosts beyond an absolute minimum, and ideally enclosed within a sandbox.

Restrict egress. Most organizational servers don’t require unrestricted connectivity to arbitrary external hosts. If implementing strict default-deny egress policies presents technical challenges, consider a streamlined approach: enforce DNS traffic resolution exclusively through your enterprise DNS server, and configure it to block known public services that are commonly abused as command-and-control infrastructure or exfiltration channels such as Pastebin or OneDrive. Similar outbound restrictions can also be enforced via an NGFW.

Avoid issuing large batches of secrets, or storing them as a single object. Even with secret managers in place, an attacker can gain broad access if a compromised account holds excessive permissions. In the Hugging Face breach, a single Kubernetes Secret object contained 136 keys. The fewer secrets issued, the lower the impact of a compromise — and the easier key-rotation becomes.

Avoid long-lived secrets and don’t store them in environment variables that are inherited by all child processes.

Isolate applications from metadata servers. Most cloud environments run services — such as the EC2 Instance Metadata Service — that issue credentials and system details upon request. Isolating these from arbitrary applications — via firewall rules or by enforcing IMDSv2 — eliminates the most direct path from a single application compromise to a full account takeover.

Audit server and application identities. Service accounts with broad access, OAuth tokens, cross-environment keys, and other artifacts granting excessive privileges are most frequently found in server and app identities rather than human user accounts.

Track usage of credentials at anomalous entry points. In the Hugging Face attack, credentials assigned to an internal host were leveraged from an external one. Detection rules that flag these anomalies are straightforward to implement and deliver high-fidelity, low-noise signals of an attack that’s already progressed beyond its initial stages.

Use short-lived, single-use keys when provisioning new devices and services. During the incident, a single key was used to register 181 separate devices without triggering a single alert.

Cybersecurity news increasingly focuses on attackers using AI for reconnaissance, vulnerability hunting, and writing scripts or malware. Against this backdrop, a less conventional category of attacks remains largely overlooked: threat actors leveraging AI tools that are already deployed within the target infrastructure.

AI agents designed for coding, as well as command-line tools — such as Claude Code CLI, Gemini CLI, Codex CLI, Amazon Q CLI, and the like — are particularly attractive targets. These agents can read and modify files, execute shell commands, install packages, and connect to external services. Developers frequently allow them to run in autonomous mode to avoid manually confirming every action. For an attacker, such an agent becomes a ready-made multitool that has the right privileges, access to critical data, and security clearances from the get-go. That said, other types of AI agents, such as automated financial workflow assistants, are equally vulnerable to exploitation.

Why trusted agents pose a threat

An AI agent differs from standard software in its broad access privileges, the wide array of ways it can be prompted to execute actions, and the difficulty of detecting its anomalous behavior. Enterprise security controls (EPP, EDR, XDR, SIEM) scan for malicious files, known command patterns, and suspicious action sequences. Even if a specific strain of malware hasn’t been seen before, its behavior often gives it away: bulk searches for documents, spawning PowerShell processes, accessing credential stores, or exfiltrating data to an external server from an unknown application will trigger a deluge of alerts in the SOC.

It’s much harder to apply that same detection logic to a legitimate, pre-installed AI agent. By design, these tools are capable of executing an extremely broad range of operations. Running scripts, reading configuration files (including secrets), moving documents, and installing dependencies can all be a part of their normal operation — especially on a developer’s machine.

When an attacker attempts to exploit this kind of AI system, they use natural language to convey their instructions to the agent. A request to hunt for passwords, API keys, or cryptocurrency wallets can be framed in thousands of ways, expressed in a variety of languages, or phrased indirectly. Furthermore, a malicious command can be hidden in any data stream the agent reads — including text fields in documents, file names, or even error logs, as shown in one of the examples below.

Integration between agents and third-party software further expands the attack surface. Agents are frequently connected to email, instant messaging apps, task management systems, knowledge bases, and MCP servers. As a result, a malicious instruction can originate from virtually any source whose data the agent processes automatically.

Below are several examples of attacks where threat actors leveraged AI agents to achieve their goals.

The Nx attack: a coding agent hunting for passwords

The most prominent and widespread example is the compromise of Nx npm packages in August 2025, dubbed s1ngularity. Attackers exploited a vulnerable GitHub Actions workflow, stole a package-publishing token, and released trojanized versions of several @nx packages that harvested developer secrets.

The malicious post-install script checked whether Claude Code, Gemini CLI, or Amazon Q CLI were installed on the victim’s machine. If any of these were detected, the agent was issued an instruction to scan for cryptocurrency wallets, ENV files, API keys, and other sensitive data. The agents were executed with auto-approval flags enabled, such as --dangerously-skip-permissions, --yolo, and --trust-all-tools.

The search results were then exfiltrated to public GitHub repositories created by the attackers. Over several waves of the attack, thousands of secrets belonging to hundreds of organizations were exposed.

Instead of developing a complex, custom harvesting tool, the attackers repurposed the victim’s own legitimate agent. It already could navigate the file system, understand file context, and identify high-value target data. However, post-incident analysis by security experts revealed drawbacks to this approach: the AI occasionally refused to carry out unethical tasks. Furthermore, the malware authors had to adjust their code so that it “waited” for the AI agent, which could take a considerable amount of time to process the task while generating significant intermediate output.

Communication via… error logs

The AgentJacking study was not a malicious attack, but it differed from one only in intent: its authors had no interest in stealing secrets. After developing the attack technique, the researchers tested it on a global scale and received callback responses from AI agents belonging to more than a hundred real-world companies — some of them major enterprises.

The attack targeted Sentry, a widely used telemetry service. Websites and mobile applications automatically transmit error reports to Sentry so developers can monitor issues without requiring active user reporting. These reports are accepted without authentication because errors often occur for anonymous visitors that the site or application knows nothing about.

The researchers configured their own MCP server within Sentry, allowing AI agents to analyze reports automatically. Next, they submitted a fake error message containing a data block formatted to match Sentry’s MCP server data structure. Embedded inside was an indirect prompt injection — an instruction to run “additional diagnostics” via a command like: npx @controlled-validation-package --diagnose.

If error handling were delegated to an agent, it could take the instruction as legitimate and execute a command resulting in the automatic installation of a third-party package. In this experiment, the package merely contacted the researchers’ command-and-control server without exfiltrating any sensitive data. A genuine malicious package, however, could steal credentials, tamper with code, or establish persistence in the system. Identifying targets for such an attack is equally straightforward: an organization’s use of Sentry is easily detectable by analyzing website code or checking the strings in mobile application binaries.

Notably, according to the study’s authors, Sentry acknowledged the issue but stopped short of implementing comprehensive fixes, opting instead to block only the specific variant of the malicious query routed through the MCP. Screenshots included in the study reveal that a Sentry employee described the system as “technically not defensible” against the attack.

Stealing financial data via MCP

Abusing unsecured MCP servers — as seen in the Sentry case — is just one way to attack this protocol. Introduced just a year ago, and already dubbed “USB for AI agents”, the Model Context Protocol standard was adopted rapidly without robust built-in security controls. As a result, attacks targeting MCP are bound to proliferate — a trend Microsoft recently warned about.

Threats can lie hidden even within the description of an MCP tool before the agent ever uses it. Agents parse tool descriptions to determine when to call the tool and what data to supply. If an MCP server provider stealthily alters a description — adding an instruction that demands collection of additional data — the agent may execute it as part of its standard workflow.

Microsoft highlights a scenario involving a financial agent that verifies vendor banking details via an external MCP service. Following a malicious update to the tool’s description, the agent begins appending unpaid-invoice data to its responses. The user sees a normal response, while confidential financial data is quietly exfiltrated to the server owner. Individually, each operation appears authorized: the agent operates under the employee’s permissions, queries an approved service, and uses a standard interface. Yet a data breach still occurs.

It remains unclear from Microsoft’s ambiguous phrasing whether this example is purely hypothetical or based on an investigation into a live incident. However, the fact that the post originated from Microsoft Incident Response points to the latter.

Prompt injections have already gone mainstream

In 2026, Unit 42 researchers identified widespread attempts to covertly embed instructions designed for AI systems across numerous public websites. Many of these instructions aim to force models to reveal system prompts, approve sites and posts during ad placement reviews, or boost phishing pages through SEO manipulation.

Similarly, an analysis of Common Crawl data by Google recorded a surge in such material: between November 2025 and February 2026, the share of malicious injections increased by 32%.

That said, confirmed cases of successful compromise using these techniques remain scarce. The presence of an injection on a webpage does not guarantee that an agent ever actually executed the command. Nevertheless, threat actors have clearly begun optimizing web content for an era when it will be read not just — and perhaps not primarily — by humans.

How to defend against agent-based attacks

The key takeaway for cybersecurity teams and all AI users alike is simple: an AI agent should never be considered trusted merely because it has been officially approved for enterprise use.

Organizations must:

• Maintain an inventory of agents, MCP servers, and the various tools available for them. Use allowlists to restrict agent tools, and package installations to components explicitly vetted and approved by information security. Allowlists must specify exact package versions.
• Monitor version changes to MCP tools and other dependencies. Re-evaluate updates before adding them to the approved inventory.
• Apply the principle of least privilege. Grant agents only the specific permissions needed for a given task, and only for the duration of that task.
• Enforce human-in-the-loop approval for package installations, script executions, file sends, and other high-risk actions.
• Run agents in isolated environments with limited access to developer workstations.
• Disable dangerous auto-approval modes.
• Restrict outbound network connections, and permit data transmission only for approved services.
• Store secrets in centralized, secure vaults, using short-lived tokens and rotating them regularly.
• Forward prompt logs, tool calls, shell commands, and agent network traffic into SIEM and XDR systems. For these tools to be effective, they require deep operational context regarding agent activity alongside specialized rule sets tailored to AI data processing pipelines.

Starting last year, Android users have been seeing the “number is now verified” system notification more frequently. And in some cases people even find mysterious outgoing text messages in their history that they never sent.

These messages often cause confusion and even anxiety: has a virus infected the phone? Are tech giants spying on our phone numbers? Let’s break down how this feature works, and what potential risks it poses to your privacy.

Why all these phone number verifications?

The notification pops up whenever Google’s phone number verification feature is triggered on your device.

Its main job is to make sure that the SIM card tied to a specific phone number is physically inside the phone. Once verified, that phone number automatically links to all active Google accounts on the device.

There are several key services that rely on the verification. Most importantly, it drives Rich Communication Services (RCS) — the modern standard for “rich” messaging right inside your standard text messaging app. It feels like a popular chat app, but without the need to install anything extra. Unlike iMessage, which is locked to Apple’s ecosystem, RCS works across compatible smartphones on various platforms since it’s an industry standard set by carriers rather than tech giants. Since recently, both Apple and Android users have been able to exchange RCS messages. For this feature to work in the Google Messages app, Google needs ongoing confirmation that your SIM card is active. If you pull out your SIM card, RCS chats will keep working for about eight days before turning off automatically.

According to Google, phone number verification serves several other purposes as well:

• Account security and recovery. A verified phone number enables quick sign-ins to your Google account, two-factor authentication, and easy password recovery.
• Emergency services and device location. This includes Find My Device, remote phone lock, and sharing your location with emergency responders — including via satellite messaging on the Pixel 9 and certain other modern smartphones.
• Better sharing on Google. This feature helps other people find you faster on Google Meet or Duo, use Quick Share to send you files, and see that the phone number is linked to your profile.

Google also recently confirmed that this data is used to counter scams. Verification helps block calls or text messages from spoofed numbers if both the real owner of the number and the recipient are using Android devices with verified phone numbers.

How does Google verify phone numbers?

Phone number verification technology has been around on Android for quite some time — Google was sending test SMS messages as far back as in 2019. However, it became widely visible to most users following a Google system update in September 2025. The process was baked deeper into the operating system, with number verification now running by default during initial phone setup, and re-running periodically in the background.

There are two main technical methods for the verification process. Which one your device uses depends on your mobile carrier and your version of Android.

The older method relies on hidden text messages. In the background, your smartphone sends a specialized technical text message to Google’s servers. The operating system intercepts this message before you ever see it, which is why it rarely appears in your standard text messaging app. However, due to software glitches or Android customization quirks, these texts occasionally surface in sent messages, startling users. They typically look like this: “(string of letters and numbers) Google is re-verifying the phone# of this device.” Google explicitly notes in its help documentation that standard messaging rates from your carrier may apply.

In recent years, direct carrier verification (via carrier APIs) has become the primary verification method. This approach is more modern and secure than previous ones. The smartphone sends an encrypted token containing device and SIM card identifiers to the mobile carrier, and the carrier responds with the confirmed phone number. The whole process takes just a couple of seconds and runs completely unnoticed by the user. Both global telecom giants and smaller providers have connected to this verification network. Notably, third-party apps can also tap into the results of this check through Firebase Phone Number Verification, getting confirmation from your mobile carrier about which phone number is active on the device.

“Other device data” and privacy concerns

In its official documentation, Google notes that device identifiers and SIM card data may be collected during verification. In practice, this refers to unique identifiers for the SIM card and its subscriber profile (ICCID and IMSI), as well as technical device identifiers needed to run mobile networks. Google also explicitly states that it does not sell your personal information, including your phone number, to anyone.

Naturally, sending additional unique identifiers to Google — especially given the scale of its advertising business — always raises concern among privacy-conscious users. Here’s what you should keep in mind:

Metadata collection. To run RCS, Google exchanges data with your mobile carrier. Even when the content of RCS messages is encrypted, metadata — such as who is messaging whom and when — can still be stored on your carrier’s servers, and in some cases, on Google’s servers. As cybersecurity experts at the Electronic Frontier Foundation point out, if privacy is your top priority, you’re better off sticking to dedicated encrypted messaging apps.

Linked accounts. If you have both a personal and a work Google account (or a personal and a family account) set up on the same phone, the verified number will automatically link to both profiles. The operating system does not offer built-in tools to separate numbers for different accounts on a single device.

Phone number leaks. Apps on your device can already access various user identifiers, including your phone number. However, this verification system makes it easier to link multiple phone numbers to a user who has multiple Google accounts. And while Google states that it never sells phone numbers, you cannot say the same with confidence about obscure third-party Android developers.

On by default. The feature is enabled out of the box, and most users have no idea their device is silently querying their carrier in the background. While you can opt out in your Google settings once you notice it, there’s no guarantee that any data already collected will actually be deleted.

Turning off verification — why and how

For most users, verification is genuinely helpful. It simplifies account recovery, makes finding a lost phone easier, powers modern text messaging features, and assists emergency services when every second counts.

However, if you want to minimize the amount of metadata sent to Google, mobile carriers, and other data brokers, you can manually disable the feature. Here’s how:

1. Open your Android smartphone’s settings.
2. Tap Google, select the account at the top, and switch to the All services tab.
3. Under Privacy and security, tap Phone number verification.
4. Turn off automatic phone number verification.

If you also want to disable Better sharing on Google, go to Settings → Google → Manage your Google Account → Personal info → Contact info → Phone, select your number, and turn off the setting.

If you have multiple Google accounts on your phone, you’ll need to repeat these steps for each one. Unfortunately, verification sometimes turns itself back on automatically, and there’s no reliable way to prevent this on standard consumer phones with stock software.

Keep in mind that disabling verification means losing access to RCS chats in Google Messages — forcing you to fall back on basic SMS, or switch to alternative secure messaging apps. You also won’t be able to use that number for quick account recovery should you forget your password.

To thoughtfully customize your privacy settings across all your devices — regardless of the operating system, browser, or app — check out our free online tool, Privacy Checker.

Curious about other privacy risks you might not even know exist? Check out our other deep dives, here:

• Post-breakup digital hygiene: what to check and shut down
• What happens in the bedroom stays in the bedroom
• Is telehealth actually safe?
• Predator vs. iPhone: the art of invisible surveillance
• Brain drain: vulnerabilities in mental health apps

When targeting an organization to steal information, maintaining a low profile is critical for attackers. They typically aim for long-term persistence, which requires avoiding security alerts while preserving access in case they’re detected and the organization initiates incident response or routine password resets. Malware such as infostealers or ostensibly legitimate remote monitoring and management (RMM) tools fail the first requirement: their use triggers EDR and generates suspicious events in SIEM consoles. Relying on stolen credentials conflicts with the second requirement: the moment the security team suspects a compromise, passwords can be changed immediately — terminating access. If attackers attempt to steal browser cookies instead of passwords, they face a different challenge: many online services now correlate device characteristics with the expected session cookie and block access if the cookie is used on an unrecognized device. Furthermore, cookie theft protection mechanisms implemented in Chromium-based browsers (such as Chrome, Edge, and Opera) this year have made this approach significantly more difficult.

To address this persistence challenge, the ToddyCat APT — whose main game is spying — developed a novel technique. Kaspersky experts discovered this method during an incident investigation and named it Shadow Token via Remote Debug (STRD). This technique allows the attackers to establish reliable, persistent access to a victim’s mailbox and other resources in Google Workspace. With minor adjustments, the same approach could potentially be adapted to target other services that grant third-party application access via OAuth 2.0 authentication.

How an STRD attack works

First, the attackers must compromise the victim’s system with malware. In past campaigns, ToddyCat gained initial access to organizations by exploiting known vulnerabilities in server software and distributing malicious loaders via messaging apps. The specific employee targeted by the attackers might not notice the intrusion at all. This can occur, for example, if the adversary first obtains privileged administrative credentials and uses them to deploy the malware onto targeted machines remotely. Crucially, the deployment and execution of this malware mustn’t trigger immediate security alerts.

Once active, the malware executes an STRD attack, connecting the attackers’ remote service to the victim’s mailbox using the OAuth 2.0 protocol. This process requires no user interaction and shows no visible activity on the screen. To the cloud environment (Google Workspace, in the case at hand), the activity appears as if the user has legitimately authorized a third-party app for email access or data backup.

After that, the malware can terminate its operations and even delete itself from the system. The adversary retains direct access to the mailbox using the acquired OAuth token — and they need no connection to the victim’s endpoint or to the corporate network for that. Depending on the organization’s Google Workspace configuration, this access can persist for an extended period and survive subsequent password resets.

The core concept of Shadow Token via Remote Debug

At the heart of this attack is a connection to Google Workspace services via OAuth 2.0. This is a legitimate workflow used whenever a third-party application requests access to calendar data, emails, or Google Drive files. For example, to display calendar meetings in Zoom and automatically generate conference links, a user must authorize Zoom to access Google services. Similarly, configuring a third-party email client or calendar app requires granting permission. During this authorization process, the service requesting access opens a new browser window. In this window, Google Workspace first prompts the user to select the appropriate account. Once the account is chosen, the subsequent screen displays the specific permissions requested by the third-party app, allowing the user to either approve or deny access. For this scenario to proceed seamlessly, the user must already be authenticated to Google services in their browser — which is typically the case in organizations using Google Workspace. If the user isn’t authenticated, additional steps for entering credentials and completing multi-factor authentication are introduced into the sequence.

The ToddyCat hackers developed a malicious tool called Umbrij to facilitate a two-step covert authorization process when the user is already authenticated in Google. First, the malware identifies the browsers installed on the system, and locates the specific folder storing the user’s active profile for each. The attackers target Chrome and Edge, as these are the browsers most likely serving as the primary ones within the organization.

Next, Umbrij copies the entire user profile folder to a different directory on the machine. It then launches an instance of the browser, specifying the path to the duplicated profile via the command line. Because this duplicate profile contains the user’s session cookies, websites with saved credentials won’t prompt for re-authentication. Furthermore, since this occurs on the exact same computer where the primary browser is running, online services detect no anomalies. The browsing history for this newly launched instance is isolated within the new folder, keeping it hidden from the user’s main account activity.

Crucially, the browser is launched in a dedicated debugging mode typically reserved for web development. The browser window and user interface don’t appear on the screen at all (headless mode). Instead, the browser can be controlled through a debugging port using the DevTools protocol, allowing the malware to issue commands and read the state of the screen. To orchestrate these actions, Umbrij leverages Puppeteer, a legitimate automation library.

After verifying that the debugging browser instance has launched successfully, Umbrij opens a legitimate Google Workspace OAuth screen within it. The request sent to Google is engineered to bypass additional security checks while requesting maximum access privileges. For the application ID — the identity supposedly requesting these extensive permissions — the malware impersonates one of two legitimate tools: Google Workspace Migration for Microsoft Outlook (GWMMO), or Google Workspace Sync for Microsoft Outlook (GWSMO).

When Google opens the window within the headless browser, Umbrij uses debugging tools to programmatically click on the corporate account name and the confirmation buttons. As a result, Google generates an authorization code for the app. Umbrij extracts and saves this code, subsequently forwarding it to the attackers’ command-and-control server. Finally, operating entirely within their own infrastructure rather than on the victim’s computer, the attackers exchange this authorization code for an OAuth access token. This single token is all they need to maintain long-term unrestricted access to the mailbox.

How to protect against OAuth token theft

If a Google Workspace account is compromised, incident response measures must include the following steps after collecting the necessary logs and other forensic data for investigation:

• Resetting the affected user’s password
• Terminating all active web sessions for the user
• Revoking OAuth tokens and third-party app permissions
• Reviewing and removing access granted through legacy App Passwords

In addition, security and IT teams must systematically audit issued OAuth permissions, revoke unjustifiable access rights, and restrict capabilities that allow excessive or unauthorized permission grants. We covered this topic in detail in our article on blocking unwanted AI assistants.

How to prevent exploitation of Shadow Token via Remote Debug

While Kaspersky users are protected against the Umbrij tool, security teams should proactively implement policies that prevent standard users from launching browsers in debugging mode. This functionality is intended exclusively for website and web app developers. This restriction can be enforced through the DeveloperToolsAvailability group policy (available for both Chrome and Edge).

Additionally, configure monitoring within your SIEM/XDR to track the launch of browser instances with an active debugging port. This event serves as a strong indicator of this specific attack technique.

Unchecked AI in the workplace quickly becomes a massive loophole for data leaks and security breaches. All too often, employees drop sensitive company data into public chatbots, or install rogue AI assistants on their own — in the process handing over way too much access. In a previous post, we broke down the different types of risky AI systems, and later shared some tips on how to turn off the built-in AI features on major tech platforms. Today let’s take a look at practical ways to block or restrict the unauthorized “helpers” employees might be using — from ChatGPT and Grammarly, to meeting bots like Fireflies and Read AI.

How to detect and restrict ChatGPT

ChatGPT is the biggest culprit when it comes to unauthorized AI use worldwide. A quick word of warning, though: an outright ban only sends users hunting for sketchy third-party sites or messaging app chatbots that hook into the same service. That’s why it’s always a good idea to offer an approved alternative before pulling the plug.

Detecting it: keep an eye on the NGFW or web filter for traffic heading to chat.openai.com, chatgpt.com, oaistatic.com, oaiusercontent.com, or cdn.oaistatic.com. It’s also smart to use EDR/EPP tools to scan browser histories, installed apps, and browser extensions across corporate devices.

Locking it down: use the firewall or web filter to block the entire AI Services category, and set up DNS to reroute traffic away from those OpenAI domains. Browser policies can also be used to ban ChatGPT-powered extensions. Better yet, block all extensions not on a pre-approved allowlist. Finally, use application controls and EPP solutions to stop users from installing the official desktop app (ChatGPT.exe or com.openai.chat).

How to detect and restrict Claude and Claude Code

Detecting it: use the NGFW or web filter to track traffic going to claude.ai, anthropic.com, *.anthropic.com, and api.anthropic.com. EDR/EPP or application control tools can also be used to scan employee computers for the desktop app (claude.exe).

Locking it down: drop a blanket block on the AI Services category through the NGFW or web filter, and tweak DNS settings to reroute traffic away from the aforementioned Anthropic domains. Next, use browser policies to shut down Claude-powered extensions. Finally, use application controls and the EPP platform to prevent users from installing the desktop app.

How to detect and restrict Perplexity AI

Detecting it: keep tabs on the NGFW or web filter to flag any traffic heading to *.perplexity.ai or pplx.ai.

Locking it down: just like the others, add the AI Services category to the NGFW or web filter blocklist, and use DNS routing to redirect traffic away from those domains.

Configure the browser to block third-party extensions from being installed. If Firefox is used in the organization, be aware that recent versions come with Perplexity built in. Luckily, these AI features can be turned-off company-wide using enterprise policies — specifically, by setting SidebarChatbot = blocked. The full list of tweaks can be found in the Firefox documentation.

How to detect and restrict DeepSeek

Detecting it: keep an eye on the NGFW or web filter for traffic hitting deepseek.com, chat.deepseek.com, api.deepseek.com, or platform.deepseek.com. For better precision, analyze the SNI (server name identification) in TLS connection requests. For mobile devices, look out for the official app (com.deepseek.chat).

Locking it down: blocklist the AI Services category on the NGFW or web filter, and reroute traffic to DeepSeek’s domains via DNS settings. Use browser policies to block third-party extensions, and lean on MDM/EMM tools to restrict the mobile app.

How to detect and restrict Mistral, xAI Grok, and Character.ai

The playbook for these tools is exactly the same as DeepSeek, so here’s the quick list of domains to watch for and block: chat.mistral.ai, mistral.ai, console.mistral.ai, grok.com, x.ai, api.x.ai, character.ai, beta.character.ai, and c.ai.

A quick word of warning on Grok: because Grok is baked into X, blocking this specific AI access point means blocking the entire social media platform.

How to detect and restrict Slack AI

Detecting it: in the Slack workspace admin dashboard, look under Analytics → Slack AI usage. If an enterprise plan is used, the detailed Slack logs can be searched for any events starting with the ai_ prefix.

Blocking it with policies: in the organization’s Slack settings, click through the Workspace settings → Roles & permissions → Feature access, and change the permission to “no one”. Slack has a step-by-step guide in their help center.

Locking it down: shutting this down at the network level is tricky; it can be pulled off with a finely tuned CASB solution in place. Also, don’t forget the importance of blocking rogue integrations and keeping external AI services from tapping into Slack data in the first place. We covered how to lock this down using OAuth controls in a previous post.

How to detect and restrict Zoom AI Companion

Detecting it: if a corporate Zoom subscription is in use, just head to Admin Center → Reports → AI Companion usage. Detecting Zoom’s AI when employees join external meetings or use free accounts is a lot tougher, but email filters can be set up to flag incoming AI-generated meeting notes by scanning for subject lines or text containing “Meeting summary” or “Meeting assets”.

Blocking it with policies: for the company’s own Zoom subscription, go to the Admin Portal → Account Management → Account Settings → Meeting → AI Companion and toggle it OFF for everyone.

Locking it down: unfortunately, AI Companion is baked into Zoom’s DNA, so the only real option is blocking Zoom altogether.

How to detect and restrict Grammarly

What looks like an innocent spellchecker is actually one of the biggest culprits for workplace data leaks.

Detecting it: check the NGFW or web filter logs for traffic hitting grammarly.com, *.grammarly.com, and gnar.grammarly.com. EDR and MDM/EMM tools can also be used to hunt down the standalone desktop apps (Grammarly Desktop.exe and the macOS version), as well as the Grammarly browser extension.

Locking it down: use firewalls to block those domains at the network level, and EPP to stop employees from installing the desktop app, browser extensions, or the Grammarly add-ins for Microsoft Word and Excel.

How to detect and restrict meeting assistants: Fireflies, Read.ai, Tactiq, Fathom, and Granola

This massive category of third-party SaaS tools records and analyzes meetings — creating a massive risk for data leaks. The trickiest part? Outside clients or vendors can bring these bots into a meeting just as easily as employees can.

Detecting them: run an audit on calendar invites, and look for bot participants using email domains like @fireflies.ai, @read.ai, @tactiq.io, @fathom.video, or @granola.ai. Zoom, Teams, or Google Meet logs can also be used to review external participants who joined past calls.

Locking them down: since it’s impossible to control what outsiders do, blocking these bots comes down to tightening meeting rules. The best moves are: blocking users from granting OAuth permissions for bots to join calls, restricting employees from inviting unapproved external participants, or locking down meeting recording access for external users. That last option is usually the least painful way to keep bots out without disrupting business.

How to detect and restrict AI code editors: Cursor, Windsurf, and the like

Detecting them: use EDR/EPP tools to scan for executables like cursor.exe or windsurf.exe. It’s also worth monitoring network traffic heading to cursor.com and windsurf.com, as well as traffic hitting various AI model API providers. Keep in mind that there’s a pretty extensive list of API hosts to monitor here, since these editors aren’t tied to just one specific AI vendor.

Blocking them with policies: these apps can be prevented from being installed by setting up filters based on the developer’s digital signature certificate. Alternatively, a strict application allowlist can be employed where only pre-approved software is allowed to run.

Locking them down: rely on the EPP/EDR platform to actively detect and block these applications from running.

How to detect and restrict local AI tools: Ollama, LM Studio, and GPT4All

On one hand, this category carries fewer data leak risks because the AI models run completely locally on the user’s machine. On the other hand, it opens up a whole new can of worms: these apps themselves aren’t always highly secure, and can become targets for cyberattacks. Plus, it still means that employees can misuse models or process data in unauthorized ways.

Detecting them: EDR/EPP tools are the best line of defense here. They should be used to flag known local AI files and processes like ollama.exe, ollama serve, lmstudio.exe, LM Studio.app, jan.exe, or gpt4all.exe. From a network perspective, it’s worth scanning for open ports on local devices — typically port 1234 for Ollama and LM Studio, or port 8080 for WebUIs (using an additional fingerprint check of the server response). Another massive red flag is the presence of large files (often several gigabytes) containing language model weights. Look out for extensions like .gguf, .bin, or sometimes .safetensors.

Locking them down: use EPP/EDR platforms or windows AppLocker to block these applications by name, or switch to an application allowlist.

How to detect and restrict autonomous agents: OpenClaw, NemoClaw, and NanoClaw

This is easily one of the most dangerous categories of AI tools out there. These agents mix high-level independence with access to untrusted data, making them a massive security headache.

Detecting them: use EPP/EDR tools to sniff out active processes like openclaw, nanoclaw, nemoclaw, or clawdbot. Also keep an eye out for devices running Node.js that suddenly start launching Bash or Python scripts. Another dead giveaway is the appearance of system folders like ~/openclaw, ~/nanoclaw, ~/.claw*, or ~/clawhub. At the network level, monitor connections to the AI model APIs we mentioned earlier, as well as traffic hitting servers like openclaw.ai, nanoclaw.dev, or clawhub.*.

Locking them down: the safest bet is to use strict application allowlisting (only allowing approved software to run), or to specifically ban the known agent apps listed above. On top of that, consider blocking non-developers from installing Node.js and Docker, neither of which they need on their computers anyway.

Lately, software developers have been baking AI features straight into everyday work tools, operating systems, and browsers. In some cases, they’re genuinely handy. However, their presence introduces specific risks, which means plenty of companies are hesitant to give employees access to these tools. In a previous post, we categorized these unwanted AI systems, looked at how to spot them at the network and endpoint levels, and covered the ultimate universal kill switch: managing OAuth access across major corporate platforms. In this deep dive, we’re getting tactical: breaking down how to disable or restrict the AI built into popular platforms.

A quick heads-up: major software vendors occasionally change the names of their AI settings and tweak how they function. If any of the options mentioned below are missing or aren’t working as expected, a quick web search for the setting’s name will usually point you to its new location or branding.

How to turn off Microsoft 365 Copilot

Detection: you can check actual Copilot usage in the logs by going to Microsoft 365 admin →  Copilot usage report.

Disabling via policies: in the Microsoft 365Admin Center, go to Settings →  Integrated Apps, find Copilot in the Available Apps list, and select Block. More granular configuration policies are available under Customization →  Policy Management. The Policies page here contains over two thousand entries, so you’ll want to filter them by the keyword “Copilot” (detailed guide). Given that Copilot is a paid add-on for Office, another way to block it — and save money by doing so — is to simply avoid assigning users SKUs that include Copilot.

We recommend separately blocking Copilot Chat, which is available in Teams, Edge, Outlook, and several other services. Yes, it’s not Copilot itself. And yes, it has to be blocked separately by following this guide.

Additional layer of protection: you can block the domains copilot.cloud.microsoft and m365.cloud.microsoft/chat at the web filter or NGFW level. However, Microsoft explicitly advises against this, warning that it could break other Microsoft 365 features.

How to turn off Windows Copilot

Beyond the Office version of Copilot, you also need to manage its consumer-facing cousin.

Detection: look through your NGFW or other network logs for traffic hitting copilot.microsoft.com, bing.com/chat, or edgeservices.bing.com.
Disabling via policies: in Windows Group Policy, navigate to Computer Config →  Admin Templates →  Windows Components →  Windows Copilot. In Microsoft 365 Group Policy, go to Admin center →  Block consumer Copilot for organizational accounts.

Additional layer of protection: block the Copilot.exe executable from running entirely.

How to turn off the Copilot sidebar in Edge

Detection: look through your NGFW or other network logs for traffic hitting copilot.microsoft.com, bing.com/chat, or edgeservices.bing.com.

Blocking: configure the following MS Edge Group Policies: HubsSidebarEnabled = false, EdgeShoppingAssistantEnabled = false, CopilotPageContext = Disabled (false), CopilotNewTabPageEnabled = false, Microsoft365CopilotChatIconEnabled = false, GenAILocalFoundationalModelSettings = 1 (note that disabling this unexpectedly requires a 1 instead of a 0).

Second layer of protection: block the domains copilot.cloud.microsoft and m365.cloud.microsoft/chat at the web filter or NGFW level. However, Microsoft explicitly advises against this, warning that it could break other features.

How to turn off the Gemini Assistant in Google Workspace

Detection: check the Workspace Admin Console (admin.google.com), Gemini usage report section.

Blocking via policies: in the Admin Console, navigate to Apps →  Additional Google services → > Gemini app, and set it to OFF. Then, go to Manage Workspace smart feature settings →  Smart features in Google Workspace, and set it to OFF.

Second layer of protection: block network traffic to the domains gemini.google.com, bard.google.com, and aistudio.google.com.

How to turn off Gemini in Google Chrome

Detection: check your Chrome Enterprise reports (Chrome management →  Reports), or look through network traffic logs for connections to the previously mentioned domains.

Blocking via policies: in your Chrome Enterprise policies, configure the following settings: GenAILocalFoundationalModelSettings = 0, HelpMeWriteSettings = 2 (disabled), TabOrganizerSettings = 2, CreateThemesSettings = 2, DevToolsGenAiSettings = 2.

Additional layer of protection: block network traffic to the domains gemini.google.com, bard.google.com, and aistudio.google.com. Additionally, block unauthorized Chrome/Chromium installations (those outside your policy management) with the help of host-based application control tools like EPP/EDR or AppLocker.

How to turn off Apple Intelligence

Detection: on your NGFW and web filters, traffic hitting apple-relay.apple.com and *.apple-cloudkit.com is a clear indicator that Apple Intelligence is active.

Blocking via policies: any managed Apple device allows you to disable individual AI features, though there isn’t a master switch you can flip to shut down “all AI”. In your MDM profile, you need to set the following keys to false (disabled): allowWritingTools, allowMailSummary, allowGenmoji, allowImagePlayground, allowImageWand, allowPersonalizedHandwritingResults, allowExternalIntelligenceIntegrations, allowExternalIntelligenceIntegrationsSignIn, allowNotesTranscription, and allowNotesTranscriptionSummary. Here is a brief configuration example:

<dict>
<key>PayloadType</key>
<string>com.apple.applicationaccess</string>
<key>allowWritingTools</key>
<false/>
<key>allowMailSummary</key>
<false/>
</dict>

Despite Apple’s shift toward declarative device management, these AI features still need to be managed through traditional MDM payload settings.

Second layer of protection: block network traffic to the hosts mentioned above — though the obvious downside for mobile devices is that this won’t work once they leave the corporate network.

While many companies are intentionally rolling out AI to boost quality and efficiency, unsanctioned AI tools are cropping up in corporate environments even faster. Software vendors are baking AI right into products companies already use (think Microsoft Copilot and Google Gemini), while employees are taking matters into their own hands and installing tools on the sly. As a result, businesses are staring down a poorly managed data leak channel: staff paste information from corporate systems into AI chatbots, sending data not just to the SaaS vendor, but straight to the developers behind the underlying AI model. Both the risks and the mitigation strategies vary depending on the type of AI system in play. We break down this broad topic, focusing heavily on tools for spotting and blocking AI at two distinct levels.

Types of unwanted AI systems

Depending on the type of AI in question, managing and blocking its use requires a different playbook. It’s essential to break down AI into four distinct categories:

• Platform-native AI capabilities. Think Microsoft Copilot, Google Gemini, and Apple Intelligence, along with AI features baked right into browsers. The tricky thing about these is that they’re built into everyday essentials, are instantly available to every user (sometimes popping up aggressively), and most importantly, vendors try to turn them on by default.
• AI companions embedded in business apps. This bucket includes Slack AI, Zoom AI Companion, Notion AI, Jira’s Rovo assistant, and the like. These are tied to a single application and are completely inseparable from it.
• Standalone web and app-based chatbots. ChatGPT, Claude, Perplexity, Character AI, local setups like LM Studio, browser extensions, and agentic browsers like Comet. Apps and services in this category are usually adopted by employees on their own without permission: classic examples of shadow AI.

• Desktop-native multi-functional agents. This group features tools like OpenClaw, NanoClaw, NemoClaw, and others. They pose the biggest threat because they come with broad access rights by default and actively process untrusted data from the open web.

How to deal with unwanted AI

Every company, depending on its industry, appetite for innovation, and risk tolerance, needs to draw its own line in the sand between recommended, approved case-by-case, and completely banned use cases for specific AI products. Regulated sectors like healthcare play by one set of rules, while retail businesses operate under an entirely different playbook. Either way, after analyzing exactly which AI tools have already slipped into the organization, corporate policies need to be fine-tuned. That’s why the first order of business is employing existing infosec and logging tools to scan corporate infrastructure.

Depending on the chosen strategy, the uncovered AI systems can be:

• Disabled or restricted by using the built-in corporate policy settings within the tools themselves
• Hard-blocked at the endpoint or network level to create a safety net against policy workarounds or configuration errors
• Transitioned to managed access, where the tool isn’t completely blocked but instead routed through a dedicated corporate gateway that checks access permissions, and monitors usage patterns

Detecting AI systems

Spotting AI requires a multi-layered approach, as different detection methods complement each other and work best against specific types of AI.

 

Technology	What it can detect
DNS	Any AI tool with an identifiable domain
Web Gateway or NGFW	Any AI tool with a recognizable request-and-response fingerprint (API endpoint paths, domains, and other indicators). Web filters can inspect traffic content, and many gateways/NGFWs now feature a dedicated category for detecting and blocking generative AI
EPP/EDR	Locally deployed LLMs (running via Ollama, LM Studio, and similar shells), native desktop apps for ChatGPT or Claude, agentic browsers, and open-source AI agents. An indirect but strong red flag is the presence of Node.js, Python, Git, Docker, or other containerization tools on machines belonging to non-technical staff
Application control	Similar to EPP/EDR, this allows to immediately block unwanted applications right out of the gate
Browser control	AI-focused browser extensions and visits to AI-themed websites. This is a lifesaver if the corporate web gateway can’t inspect encrypted traffic
SaaS Security Posture Management (SSPM) / Identity Governance	OAuth permissions requested by AI apps and services, as well as any third-party integrations plugging into core productivity hubs (Microsoft 365, Google Workspace, and others)

 

Naturally, almost all of these tools allow to do more than just spot AI — they let to block it entirely, or at the very least, sound the alarm for the team in charge.

Keeping an eye on OAuth

Popular office AI solutions — especially meeting assistants, email and calendar automation agents, and the like — gain access to corporate data by requesting OAuth permissions directly from communication, document workflow, or video conferencing platforms. If a user has the green light to grant these permissions to third-party apps, the resulting data leaks completely bypass the organization’s perimeter. Tools like EDR and NGFW won’t see a thing when a tool like Read.ai grabs recordings of every single meeting in, say, Microsoft Teams.

The most drastic — and often best — move is to block standard users from granting OAuth consent in the first place. Here’s how to handle the technical heavy lifting (Global Administrator, Application Administrator, or equivalent rights are needed):

Microsoft 365 / Entra ID

In the Microsoft Entra admin center, head over to Identity > Applications > Enterprise apps > Consent and permissions > User consent settings. There User consent for applications can be disabled (check out Microsoft’s full guide).

Google Workspace

In the Google Admin console, navigate to Security > Access and data control > API controls. Under Manage App Access, the trust level for all apps can be set: Trusted, Limited, Specific Google data, or Blocked. However, the real kicker here is the Unconfigured app settings subsection, which dictates what happens when a user tries to connect an unknown app. To seal this loophole, select Don’t allow users to access any third-party apps.

A separate subsection, Manage Google Services, permits fine-tuning exactly how third-party apps interact with Google Workspace and Google Cloud services. This allows to cut off access for each individual Google product (see Google’s official guide).

Salesforce

In Setup, use the Quick Find box to search for connected apps, then select Manage Connected Apps from the results. While settings are configured for each external app individually, all users can approve access by default. There isn’t a blanket block switch here; instead, Salesforce allows to opt for Admin approved users are pre-authorized (see the full Salesforce guide on this).

Slack

From the Admin settings menu, head to Apps and workflows -> App Management Settings. Tweak the Require approved apps setting by selecting Only allow pre-approved apps. Once that’s locked in, double-check that no rogue AI tools have slipped onto the approved list.

AI security covers more than just data theft prevention, restricting rogue AI agents, or stopping assistants from giving harmful advice. A relatively simple but rapidly scaling threat has emerged: attempts to hijack computational power and exploit someone else’s neural network for personal gain. This is known as LLMjacking. With AI compute costs widely predicted to surge dramatically, the number of attackers driven by these motives is poised to grow. Consequently, when deploying proprietary AI servers and their supporting ecosystems like RAG or MCP, it’s critical to establish rigorous security measures from day one.

Statistics from a honeypot

The speed and scale of these resource-hijacking attempts are best illustrated by an experiment documented in detail in April 2026. The investigator configured a Raspberry Pi to masquerade as a high-performance private AI server, and made it accessible from the internet. When queried, it reported the availability of Ollama, LM Studio, AutoGPT, LangServe, and text-gen-webui servers — all tools commonly used as wrappers for locally hosted AI models. The server also appeared ready to accept API requests in the OpenAI format, which has become the industry standard.

All these services were seemingly powered by a local instance of Qwen3-Coder 30B Heretic, one of the most powerful open-source models, with its safety alignment removed. To throw in a sweetener, the honeypot reported the presence of various RAG databases and an MCP server with tempting capabilities like get_credentials on board.

In reality, the Raspberry Pi was simply hosting 500 pre-saved responses from an actual Qwen3 model, with a lightweight script selecting the most relevant answer for each incoming query. This setup was enough to pass a superficial check while allowing the researcher to probe the attackers’ intentions.

According to the author, Shodan, a popular internet scanning service, discovered the server within three hours of its going live. Just one hour later, requests resembling capability reconnaissance began pouring in. Over the following month, the server handled more than 113 000 requests from thousands of unique IPs, with 23% of that traffic specifically targeted at discovering AI capabilities and exploiting local LLMs and AI agents.

Requests to endpoints like /api/tags and /v1/models allow attackers to fingerprint which models are hosted on a server, while scanning for /.cursor/rules typically precedes an attempt to exploit an AI agent. Similarly, checking /.well-known/mcp.json serves as an inventory of the victim’s MCP servers. While the author makes no mention of the total number of attacks that progressed beyond simple scanning, there were 175 active attempts to hijack the LLM during the final week of the experiment alone.

What are the attackers after?

Based on the researcher’s observations, none of those targeting the decoy server attempted to execute arbitrary code or gain root access. (Editorial note: this is surprising and may point to gaps in logging.) Almost all attacks were aimed at siphoning resources. For example, the following activities were logged during the experiment:

• A well-structured attempt to parse technical documentation for a microprocessor
• A prompt to write an erotic novel
• Requests to parse and structure social media text data regarding new vulnerabilities
• An attempt to call Anthropic models using the compromised server as an API proxy

It’s worth noting that the reconnaissance of AI resources uses standardized and rapidly evolving tools. Requests from an application named LLM-Scanner originated from the infrastructure of seven different cloud providers across eight countries, suggesting that the raiders have put established methodologies in place, as well as specialized platforms for sharing techniques. By the third week of the experiment, the scanner had been updated with an additional check: it now used simple abstract questions to determine whether it’s interacting with live AI or a honeypot returning canned responses.

Among the non-specific attacks, the experiment recorded numerous attempts to exfiltrate credentials from the .env file. Attackers systematically hunted for this file across every conceivable directory on the server. Leaving an .env file publicly accessible is one of the most elementary mistakes when deploying projects on Laravel, Node.js, and other frameworks, yet it remains a common oversight — particularly among beginners and vibe coders. Consequently, attackers have every reason to expect their efforts to pay off.

Conclusions and defense tips

Scanning publicly accessible servers and attempting to exploit them is nothing new, but the rise of LLMs gives attackers another way to monetize their efforts — one that’s both highly lucrative for them and devastating for their victims. To understand how massive these attacks could become, look at their closest counterpart: the cryptojacking market — where criminals mine cryptocurrency using stolen computational resources. That market grew by 20% in 2025 alone. As AI-powered solutions proliferate, and as major providers hike subscription costs while local AI chips remain in short supply, we should expect LLMjacking to become an industrial-scale phenomenon.

Key defensive measures for private AI infrastructure

• For AI systems running locally on a single machine, ensure that servers like LM Studio, Ollama, or similar are configured to accept connections only on the local interface (localhost), rather than all available network interfaces. This restricts LLM access to the host machine itself, and prevents the AI from being reachable over the internet.
• For servers handling remote requests — even if the server only operates within a local corporate network — implement robust authentication and authorization rather than relying solely on API key validation. Solutions based on OIDC or OAuth2 with short-lived tokens are the most effective. This not only defends against LLMjacking, but also allows for more granular tracking of user activity, and prevents API key abuse. Furthermore, keys must be protected from more than just external attackers; a growing risk is the misuse of keys by AI agents themselves. This applies to LLM interfaces as well as MCP, RAG, and others.
• Use network segmentation and IP allowlists to give AI server access only to the departments, employees, and services that require it.
• Ensure that all client-server connections are secured with a current version of TLS.
• Apply the principle of least privilege by separating access to specific services; for instance, MCP and LLM components should have their own distinct access tokens.
• Ensure an EDR security agent is installed on all workstations and servers, including those hosting AI models.
• Monitor AI resource consumption, establish usage quotas for different employee roles, and set up alerts for anomalous activity spikes.
• Maintain detailed logs of LLM responses and requests made to the model and its supporting tools. Integrate these data sources with your SIEM. Ensure logs are resilient against tampering or deletion.

It’s best to think of the modern car as a computer on wheels — one that constantly offloads diagnostic data to the manufacturer or dealer’s servers. On board, you’ll find dozens of sensors: everything from GPS, speedometers, and hands-free microphones, to external cameras and the less obvious (but highly active) sensors for pedal pressure, tire pressure, engine temperature, and more. Even if this data isn’t beamed to the manufacturer in real-time, it’s logged in the car’s internal memory, and can reveal a wealth of information about a driver’s trips, habits, and surroundings. We’ve already taken a deep dive into how automakers collect data for commercial use, and who they sell it to (spoiler alert: insurance companies are the biggest buyers of telemetry), but today we’re looking at how law enforcement and intelligence agencies tap into this goldmine.

Digital evidence

Police departments across the globe have recognized the immense value of data stored within vehicles. If a car or its owner is potentially linked to a crime, investigators do more than just check for prints or DNA. Car Intelligence (CARINT) technology allows them to essentially scour all onboard computers, extracting data such as:

• GPS-based trip history
• Call logs, media player activity, and voice commands
• Lists of paired devices and synced contact lists
• Driving statistics: mileage, engine performance modes, and other technical parameters

There are numerous precedents where this data has served as evidence and dismantled alibis. In one U.S. criminal case, a recorded voice command became a smoking gun, proving the suspect was behind the wheel of a stolen vehicle.

With the rise of connected cars equipped with their own SIM cards and direct links to the manufacturer, law enforcement no longer needs physical access to the vehicle. Key data, such as GPS location history, can be pulled directly from the manufacturer’s servers. Furthermore, a U.S. Senate investigation revealed that nine out of 14 surveyed automakers were providing this data without a warrant.

Major suppliers of car intelligence software, such as Ateros, Berla, TA9/Rayzone, and Toka, sell their solutions exclusively to government and law enforcement agencies, which is why they’ve remained largely out of the public eye.

Comprehensive surveillance

To track persons of interest, data pulled from the vehicle itself is cross-referenced with information from other sources. According to media leaks, flagship products in this category aggregate data from the car’s SIM card, Bluetooth communication trails, street-level CCTV footage, and commercially available information from data brokers. This hybrid dataset simplifies the comprehensive mapping of a target’s movements and contacts. Journalists have discovered that some companies even market the ability to activate a vehicle’s microphones and cameras remotely and covertly, enabling real-time eavesdropping on conversations. However, experts note that due to the diversity of technical implementations across different systems, hacking the car itself remains a difficult task with no sure way of succeeding. Often, it’s simpler to correlate other, more accessible datasets to achieve the same result.

Factory-installed spy tools

Features like covert activation of cameras, microphones, and other sensors may theoretically be part of a vehicle’s stock functionality rather than the result of a hack. While we haven’t found any public evidence of such cases, it’s well known that Chinese-made vehicles are coming under increased scrutiny in several countries. For instance, they’ve been banned from Israeli military sites — with the exception of a single Chery model, provided its multimedia system is removed. Similar bans exist in the UK and Poland; furthermore, UK Ministry of Defense employees are instructed not to connect their work phones to Chinese-made cars. In Germany, security analyses of Chinese vehicles were conducted by the specialized agencies BfV and ZITiS, but the findings remain classified.

Low-cost surveillance

Tracking a vehicle — or even thousands of them — doesn’t necessarily require hacking onboard systems or tapping into vast networks of license plate readers. A recent scientific study demonstrated that innocent tire pressure monitoring systems (TPMS) provide enough data for effective tracking. Data from these sensors is transmitted via radio without any encryption and includes a unique ID that makes identifying a specific car easy. This allows for more than just confirming the vehicle’s movement; it can even be used to estimate the driver’s weight or determine if they are traveling alone. While this might not sound as impressive as remotely accessing a car’s cameras, it requires very little financial investment and works even on relatively old vehicles without an internet connection.

What you can do about vehicle tracking

While tracking a person through their car is undoubtedly a privacy risk, striking a balance in mitigating this threat is difficult: many measures are complex, largely ineffective, and simultaneously reduce the utility, safety, and convenience of a modern vehicle. Consequently, any steps taken should be weighed against your personal risk profile.

To reduce the risk of data leaks, check the privacy settings in the manufacturer’s app, the car’s infotainment system, and your connected smartphone. A connected car can transmit data about its operation to the cloud: information about trips, location, driving style, vehicle condition, and the operation of its components. Some of this data is necessary for navigation, diagnostics, and service, but not all permissions are required — check your settings and disable the transmission of data not related to the functions you need.

Be careful with permissions for access to the microphone, camera, contacts, messages, and geolocation. Only connect your own devices to the car and don’t save other people’s phones or unfamiliar Bluetooth devices in the system. When syncing your smartphone, select only the features you need — such as calls, music, and navigation — rather than granting full access to all your phone’s data.

Do not use the services of technicians who offer to “unlock” your car, reflash electronic control units, or install unofficial software to expand features, increase power, or otherwise interfere with the car’s operation. Such software has not been tested by the manufacturer: it may behave unpredictably, collect and transmit your data to malicious actors, disable security features, or affect critical vehicle systems — including steering, braking, or engine operation.

And when choosing a new car, ask the dealer not only about the number of stars in NCAP safety tests, engine power, or fuel economy, but also about the cybersecurity technologies used in the vehicle. Solutions such as the Kaspersky Automotive Secure Gateway, based on KasperskyOS, will provide the necessary protection for new cars against cyberthreats.

What other threats do connected cars hide? Read more in our posts:

• Highway to… hacked: cyberthreats to connected cars
• Car hacking via Bluetooth
• Botnets on wheels: the mass hacking of dashcams
• How millions of Kia cars could be tracked
• I know how you drove last summer
• Spies on wheels: how carmakers collect and then resell information

The entry barriers for app development have plummeted in recent times — with nearly anyone now able to build a professional website, personal news bot, or dashboard simply by giving a chatbot or AI agent a few instructions in natural English. Unfortunately, a massive gap exists between a slick prototype and a reliable, production-ready, secure application. To avoid becoming the subject of another AI fail story, or losing money and sensitive data, follow these straightforward tips. These are intended specifically for non-technical creators and very small teams. Larger enterprises should follow more sophisticated recommendations.

The primary risks of AI-generated code

While vibe coding can deliver a seemingly functional app in just a few hours, it will likely contain dangerous flaws. AI models are trained on code samples from across the internet, which often include suboptimal tutorials, buggy snippets, and outright junk. Sometimes this code simply fails to run, but more often the situation is subtler and more hazardous: the app appears to work, yet under the hood, it might rely on a crude imitation of the required logic or contain critical vulnerabilities. According to a study by the Cloud Security Alliance AI Safety Initiative, the following facts should be considered when using AI for coding:

• At least 45% of AI-generated code contains dangerous vulnerabilities, such as failing to verify the user before granting access to sensitive data.
• A professional developer using AI can write code three to four times faster, but may introduce 10 times as many vulnerabilities.
• Twenty percent of AI-generated code attempts to use external libraries and modules that don’t actually exist.
• Even when an application handles confidential data — such as payments, private messages, or documents — AI-generated code sometimes skips credential verification entirely. This can leave the app’s data open for anyone on the internet to read.
• In other instances, the app might correctly prompt for a username and password but fail to enforce access controls, allowing any registered user to view everyone else’s data.
• Access keys (tokens) for databases and AI services may be embedded directly into the source code, easy to steal, and difficult to rotate after a data breach or cyberattack.
• Project code or critical build outputs are often deployed to servers without proper access restrictions, leaving both the application logic and sensitive access keys vulnerable to theft.
• AI may implement insecure database access patterns, which can allow attackers to bypass the application to steal data or execute arbitrary code on the database server.
• Apps that include API functionality often suffer from insecure API implementations, lacking both user permission checks and rate limiting.

Core principles of securing vibe code

Always verify. Treat AI-generated code as a rough draft. It should always be reviewed and rigorously tested. Ideally, professional developers should handle this; however, if none are available, the vibe-coder should at least test the application themselves, have friends or colleagues poke around the live app, and ask them to review key code snippets. It’s also possible to evaluate code integrity by submitting a separate prompt to the AI: “Review this code for secure development best practices and check for OWASP Top 10 vulnerabilities”.

Protect secrets. Never include passwords, API keys, or any other sensitive data in AI prompts. Instead, instruct the AI to write code that securely stores all secrets in environment variables (special hidden settings).

Prioritize efforts. The main risks emerge when an application is network-accessible to outsiders, processes valuable data, or runs on infrastructure that would be useful to attackers. The components of an app or system that meet these criteria are precisely what’s needed to be protected first. A static website composed of three HTML pages faces significantly lower risk than a loyalty program integrated into an online store.

Make security an explicit requirement. Even a simple, straightforward line in the prompt, like “Follow industry standards and security best practices when generating this code”, improves the output. Providing more specific requirements for critical code snippets makes the results even better.

Don’t trust default settings. Often, the danger in vibe coding lies in the configuration rather than the code itself. For example, an app processing sensitive company data might be deployed on a public vibe-coding platform (Lovable or the like), and remain accessible to the entire internet by default. Even if the code is flawless, making that information public is a critical security failure. Because of this, every component — from hosting and database settings to the deployment pipeline — must be manually reviewed and properly configured. If the purpose of a setting is unclear, consult a chatbot for the optimal values, specifying that its goal is to enhance security, and describing who the app is intended for.

Security is a continuous process. Securing the app should not be treated as a one-off task. Every time an application is updated, hosting providers are changed, or a project undergoes any other major shift, all steps in making it secure should be revisited, and the risks reassessed.

Tips for securing vibe code

It’s natural to want an app built from broad prompts like “Make me a beautiful, user-friendly, fast, reliable, and secure app for [use case].” However, for the results to actually be effective, each of those requirements needs to be fleshed out. Below, we’ve outlined recommendations for building standard components that will make vibe code more secure. It’s important to emphasize that “more secure” doesn’t mean “perfectly secure” — these approaches lower the risk, but that risk remains well above zero.

Demand security from the AI. When assigning a task to a neural network, be explicit: “write secure code, validate data, encrypt passwords”. Each type of task requires its own security prompt. For instance, don’t just ask to “build a login form”. Instead, ask for a “secure login form with credential validation, authentication and authorization (user permissions) controls, brute-force protection, password hashing according to modern standards, transmission strictly over HTTPS, and no hardcoded secrets”. It makes sense to use these secure requirement templates every time. It’s also helpful to keep a short cheat sheet of standard requirements for AI prompts: “validate all external data and user input before processing”, “no secrets in code”, “protect APIs from abuse”, “restrict user permissions”, and “secure default settings”.

Use off-the-shelf solutions. If an app needs a user management system, insist on using a popular, reputable library, such as NextAuth, Auth0, and so on, rather than inventing a new and vulnerable solution. This is the most common cause of data breaches. This applies to more than just login and registration; for other high-risk actions like file uploads and API call processing, it’s better to use established frameworks and libraries with built-in protections rather than building everything from scratch.

Don’t trust the AI blindly; verify open-source components. Neural networks often try to inject non-existent components and libraries into a project or suggest outdated versions. Always search for the suggested names online to ensure they are real, widely used, and secure — and make sure the latest versions are used.

Demand robust encryption. Explicitly state that modern industry standards must be used for both data transmission and storage: TLS 1.3 based on OpenSSL for network traffic; argon2 or bcrypt for hashing credentials; and so on.

Never trust user input. Always instruct the AI to include validation for any data entered by users, whether in forms or search bars. Use terms like “parameterization” and “sanitization” to emphasize that the app needs protection against malicious actors, not just users’ typos.

Set limits on user actions. Require the AI to implement rate limiting for login attempts or general requests. This will protect a project from automated attacks like DoS and brute-force password guessing.

Hide the system’s inner workings. If the site crashes, users should see a simple apology page rather than a detailed error report containing snippets of the code. That kind of information is a goldmine for hackers.

Remember that you’re a developer, and you need to protect development-related digital assets. All related accounts — such as access to GitHub, project hosting, and other resources — are prime targets for attackers. Be sure to enable two-factor authentication (2FA) on all work accounts.

Make backups. Regularly back up a project both locally and to the cloud to protect it against critical AI errors as well as cyberattacks. These backups should include both the application’s source code and its databases.

Set up a sandbox. Test new features and app versions in a secure environment using a clone of an active site or app and a copy of a database. Always run thorough tests before pushing an update live. This allows catching issues without putting users or their data at risk.

Update dependencies and scan them for vulnerabilities. A vibe-coded app will almost certainly rely on third-party libraries and components, known as dependencies. It’s wise to update these regularly by rebuilding an app with the latest versions, even if app’s code itself has not been changed. This process helps patch known security flaws in the used packages.

Check for secrets leaking into the repository. Use secrets scanners like TruffleHog to audit resulting code. Even with instructions, AI might slip up and include an API key or password in the source code. A scanner ensures that files containing keys and passwords don’t end up in Git or get published alongside the project.

At the NDSS Symposium 2026 in San Diego in February, a group of respected researchers presented a study unveiling the AirSnitch attack, which bypasses the Wi-Fi client isolation feature — also commonly known as guest network or device isolation. This attack allows connecting to a single wireless network via an access point, and then gaining access to other connected devices, including those using entirely different service set identifiers (SSIDs) on that same hardware. Targeted devices could easily be running on wireless subnets protected by WPA2 or WPA3 protocols. The attack doesn’t actually break encryption; instead, it exploits the way access points handle group keys and packet routing.

In practical terms, this means that a guest network provides very little in the way of real security. If your guest and employee networks are running on the same physical device, AirSnitch allows a connected attacker to inject malicious traffic into neighboring SSIDs. In some cases, they can even pull off a full-blown man-in-the-middle (MitM) attack.

Wi-Fi security and the role of isolation

Wi-Fi security is constantly evolving; every time a practical attack is made against the latest generation of protection, the industry shifts toward more complex algorithms and procedures. This cycle started with the FMS attacks used to crack WEP encryption keys, and continues to this day: recent examples include the KRACK attacks on WPA2, and the FragAttacks, which impacted every security protocol version from WEP all the way through WPA3.

Attacking modern Wi-Fi networks effectively (and quietly) is no small feat. Most professionals agree that using WPA2/WPA3 with complex keys and separating networks based on their purpose is usually enough for protection. However, only specialists really know that client isolation was never actually standardized within the IEEE 802.11 protocols. Different manufacturers implement isolation in completely different ways — using Layer 2 or Layer 3 of network architecture; in other words, handling it at either the router or the Wi-Fi controller level — meaning the behavior of isolated subnets varies wildly depending on your specific access point or router model.

While marketing claims that client isolation is perfect for keeping restaurant or hotel guests from attacking one another — or ensuring corporate visitors can’t access anything but the internet — in reality, isolation often relies on people not trying to hack it. This is exactly what the AirSnitch research highlights.

Types of AirSnitch attacks

The name AirSnitch doesn’t just refer to a single vulnerability, but a whole family of architectural flaws found in Wi-Fi access points. It’s also the name of an open-source tool used to test routers for these specific weaknesses. However, security professionals need to keep in mind that there’s only a very thin line between testing and attacking.

The model for all these attacks is the same: a malicious client is connected to an access point (AP) where isolation is active. Other users — the targets — are connected to the same SSID or even different SSIDs on that same AP. This is a very realistic scenario; for example, a guest network might be open and unencrypted, or an attacker could simply get the guest Wi-Fi password by posing as a legitimate visitor.

For certain AirSnitch attacks, the attacker needs to know the victim’s MAC or IP address beforehand.  Ultimately, how effective each attack is depends on the specific hardware manufacturer (more on that below).

GTK attack

After the WPA2/WPA3 handshake, the access point and the clients agree on a Group Transient Key (GTK) to handle broadcast traffic. In this scenario, the attacker wraps packets destined for a specific victim inside a broadcast traffic envelope. They then send these directly to the victim while spoofing the access point’s MAC address. This attack only allows for traffic injection, meaning the attacker won’t receive a response. However, even that is enough to deliver malicious ICMPv6 routing advertisements, or DNS and ARP messages to the client — effectively bypassing isolation. This is the most universal version of the attack working on any WPA2/WPA3 network that uses a shared GTK. That said, some enterprise-grade access points support GTK randomization for each individual client, which renders this specific method ineffective.

Broadcast packet redirection

This version of the attack doesn’t even require the attacker to authenticate at the access point first. The attacker sends packets to the AP with a broadcast destination address (FF:FF:FF:FF:FF:FF) and the ToDS flag set to 1.  As a result, many access points treat this packet as legitimate broadcast traffic; they encrypt it using the GTK, and blast it out to every client on the subnet, including the victim. Just like in the previous method, traffic specifically meant for a single victim can be pre-packaged inside.

Router redirection

This attack exploits an architectural gap between Layer 2 and Layer 3 security found in some manufacturers’ hardware. The attacker sends a packet to the access point, setting the victim’s IP address as the destination at the network layer (L3).  However, at the wireless layer (L2), the destination is set to the access point’s own MAC address, so the isolation filter doesn’t trip. The routing subsystem (L3) then dutifully routes the packet back out to the victim, bypassing the L2 isolation entirely. Like the previous methods, this is another transmit-only attack where the attacker can’t see the reply.

Port stealing to intercept packets

The attacker connects to the network using a spoofed version of the victim’s MAC address, and floods the network with ARP responses claiming, “this MAC address is on my port and SSID”.  The target network’s router updates its MAC tables, and starts sending the victim’s traffic to this new port instead. Consequently, traffic intended for the victim ends up with the attacker — even if the victim is connected to a completely different SSID.

In a scenario where the attacker connects via an open, unencrypted network, this means traffic meant for a client on a WPA2/WPA3-secured network is actually broadcast over the open air, where not only the attacker but anyone nearby can sniff it.

Port stealing to send packets

In this version, the attacker connects directly to the victim’s Wi-Fi adapter, and bombards it with ARP requests spoofing the access point’s MAC address. As a result, the victim’s computer starts sending its outgoing traffic to the attacker instead of the network. By running both stealing attacks simultaneously, an attacker can, in several scenarios, execute a full MitM attack.

Practical consequences of AirSnitch attacks

By combining several of the techniques described above, a hacker can pull off some pretty serious moves:

• Complete bidirectional traffic interception for a MitM attack. This means they can snatch and modify data moving between the victim and the access point without the victim ever knowing.
• Hopping between SSIDs. An attacker sitting on a guest network can reach hosts on a locked-down corporate network if both are running off the same physical access point.
• Attacks on RADIUS. Since many companies use RADIUS authentication for their corporate Wi-Fi, an attacker can spoof the access point’s MAC address to intercept initial RADIUS authentication packets. From there, they can brute-force the shared secret. Once they have that, they can spin up a rogue RADIUS server and access point to hijack data from any device that connects to it.
• Exposing unencrypted data from “secure” subnets: Traffic that’s supposed to be sent to a client under the protection of WPA2/WPA3 can be retransmitted onto an open guest network, where it’s essentially broadcast for anyone to hear.

To pull off these attacks effectively, a hacker needs a device capable of simultaneous data transmission and reception with both the victim’s adapter and the access point. In a real-world scenario, this usually means a laptop with two Wi-Fi adapters running specifically configured Linux drivers. It’s worth noting that the attack isn’t exactly silent: it requires a flood of ARP packets, it can cause brief Wi-Fi glitches when it starts, and network speeds might tank to around 10Mbps. Despite these red flags, it’s still very much a practical threat in many environments.

Vulnerable devices

As part of the study, several enterprise and home access points and routers were put to the test. The list included products from Cisco, Netgear, Ubiquiti, Tenda, D-Link, TP-Link, LANCOM, and ASUS, as well as routers running popular community firmware like DD-WRT and OpenWrt. Every single device tested was vulnerable to at least some of the attacks described here. Even more concerning, the D-Link DIR-3040 and LANCOM LX-6500 were susceptible to every single variation of AirSnitch.

Interestingly, some routers were equipped with protective mechanisms that blocked the attacks, even though the underlying architectural flaws were still present. For example, the Tenda RX2 Pro automatically disconnects any client whose MAC address appears on two BSSIDs simultaneously, which effectively shuts down port stealing.

The researchers emphasize that any network administrator or IT security team serious about defense should test their own specific configurations. That’s the only way to pinpoint exactly which threats are relevant to your organization’s setup.

How to protect your corporate network from AirSnitch

The threat is most immediate for organizations running guest and corporate Wi-Fi networks on the same access points without additional VLAN segmentation. There are also significant risks for companies using RADIUS with outdated settings or weak shared secrets for wireless authentication.

The bottom line is that we need to stop viewing client isolation on an access point as a real security measure, and start seeing it as just a convenience feature. Real security needs to be handled differently:

• Segment the network using VLANs. Each SSID should have its own VLAN, with strict 802.1Q packet tagging maintained all the way from the access point to the firewall or router.
• Implement stricter packet inspection at the routing level — depending on the hardware capabilities. Features like Dynamic ARP Inspection, DHCP snooping, and limiting the number of MAC addresses per port help defend against IP/MAC spoofing.
• Enable individual GTK keys for each client, if your equipment supports it.
• Use more resilient RADIUS and 802.1X settings, including modern cipher suites and robust shared secrets.
• Log and analyze EAP/RADIUS authentication anomalies in your SIEM. This helps track many attack attempts beyond just AirSnitch. Other red flag events to watch for include the same MAC address appearing on different SSIDs, spikes in ARP requests, or clients rapidly jumping between BSSIDs or VLANs.
• Apply security at higher levels of the network topology. Many of these attacks lose their punch if the organization has universally implemented TLS and HSTS for all business application traffic, requires an active VPN for all Wi-Fi connections, or has fully embraced a Zero Trust architecture.

If you don’t go searching for AI services, they’ll find you all the same. Every major tech company feels a moral obligation not just to develop an AI assistant, integrated chatbot, or autonomous agent, but to bake it into their existing mainstream products and forcibly activate it for tens of millions of users. Here are just a few examples from the last six months:

• Microsoft is forcibly turning compatible Windows PCs into “AI PCs”, and automatically installing and activating Copilot for anyone with Microsoft 365 desktop apps.
• Google activated Gemini for all U.S. Chrome users, cranked its browser functionality to the max, aggressively expanded the reach of AI Overviews in search results, and baked a whole suite of AI features into its online services (Gmail, Google Docs, and others).
• Apple integrated its own Apple Intelligence (conveniently sharing the AI acronym) into the latest OS versions across all device types and most of its native apps.
• Meta added AI translations and a Meta AI chat to WhatsApp, while simultaneously banning third-party chatbots in the messaging app as of January 15, 2026.

On the flip side, geeks have rushed to build their own “personal Jarvises” by renting VPS instances or hoarding Mac minis to run the OpenClaw AI agent. Unfortunately, OpenClaw’s security issues with default settings turned out to be so massive that it’s already been dubbed the biggest cybersecurity threat of 2026.

Beyond the sheer annoyance of having something shoved down your throat, this AI epidemic brings some very real practical risks and headaches. AI assistants hoover up every bit of data they can get their hands on, parsing the context of the websites you visit, analyzing your saved documents, reading through your chats, and so on. This gives AI companies an unprecedentedly intimate look into every user’s life.

A leak of this data during a cyberattack — whether from the AI provider’s servers or from the cache on your own machine — could be catastrophic. These assistants can see and cache everything you can, including data usually tucked behind multiple layers of security: banking info, medical diagnoses, private messages, and other sensitive intel. We took a deep dive into how this plays out when we broke down the issues with the AI-powered Copilot+ Recall system, which Microsoft also planned to force-feed to everyone. On top of that, AI can be a total resource hog, eating up RAM, GPU cycles, and storage, which often leads to a noticeable hit to system performance.

For those who want to sit out the AI storm and avoid these half-baked, rushed-to-market neural network assistants, we’ve put together a quick guide on how to kill the AI in popular apps and services.

How to disable AI in Google Docs, Gmail, and Google Workspace

Google’s AI assistant features in Mail and Docs are lumped together under the umbrella of “smart features”. In addition to the large language model, this includes various minor conveniences, like automatically adding meetings to your calendar when you receive an invite in Gmail. Unfortunately, it’s an all-or-nothing deal: you have to disable all of the “smart features” to get rid of the AI.

To do this, open Gmail, click the Settings (gear) icon, and then select See all settings. On the General tab, scroll down to Google Workspace smart features. Click Manage Workspace smart feature settings and toggle off two options: Smart features in Google Workspace and Smart features in other Google products. We also recommend unchecking the box next to Turn on smart features in Gmail, Chat, and Meet on the same general settings tab. You’ll need to restart your Google apps afterward (which usually happens automatically).

How to disable AI Overviews in Google Search

You can kill off AI Overviews in search results on both desktops and smartphones (including iPhones), and the fix is the same across the board. The simplest way to bypass the AI overview on a case-by-case basis is to append -ai to your search query — for example, how to make pizza -ai. Unfortunately, this method occasionally glitches, causing Google to abruptly claim it found absolutely nothing for your request.

If that happens, you can achieve the same result by switching the search results page to Web mode. To do this, select the Web filter immediately below the search bar — you’ll often find it tucked away under the More button.

A more radical solution is to jump ship to a different search engine entirely. For instance, DuckDuckGo not only tracks users less and shows little ads, but it also offers a dedicated AI-free search — just bookmark the search page at noai.duckduckgo.com.

How to disable AI features in Chrome

Chrome currently has two types of AI features baked in. The first communicates with Google’s servers and handles things like the smart assistant, an autonomous browsing AI agent, and smart search. The second handles locally more utility-based tasks, such as identifying phishing pages or grouping browser tabs. The first group of settings is labeled AI mode, while the second contains the term Gemini Nano.

To disable them, type chrome://flags into the address bar and hit Enter. You’ll see a list of system flags and a search bar; type “AI” into that search bar. This will filter the massive list down to about a dozen AI features (and a few other settings where those letters just happen to appear in a longer word). The second search term you’ll need in this window is “Gemini“.

After reviewing the options, you can disable the unwanted AI features — or just turn them all off — but the bare minimum should include:

• AI Mode Omnibox entrypoint
• AI Entrypoint Disabled on User Input
• Omnibox Allow AI Mode Matches
• Prompt API for Gemini Nano
• Prompt API for Gemini Nano with Multimodal Input

Set all of these to Disabled.

How to disable AI features in Firefox

While Firefox doesn’t have its own built-in chatbots and hasn’t (yet) tried to force upon users agent-based features, the browser does come equipped with smart-tab grouping, a sidebar for chatbots, and a few other perks. Generally, AI in Firefox is much less “in your face” than in Chrome or Edge. But if you still want to pull the plug, you’ve two ways to do it.

The first method is available in recent Firefox releases — starting with version 148, a dedicated AI Controls section appeared in the browser settings, though the controls are currently a bit sparse. You can use a single toggle to completely Block AI enhancements, shutting down AI features entirely. You can also specify whether you want to use On-device AI by downloading small local models (currently just for translations) and configure AI chatbot providers in sidebar, choosing between Anthropic Claude, ChatGPT, Copilot, Google Gemini, and Le Chat Mistral.

The second path — for older versions of Firefox — requires a trip into the hidden system settings. Type about:config into the address bar, hit Enter, and click the button to confirm that you accept the risk of poking around under the hood.

A massive list of settings will appear along with a search bar. Type “ML” to filter for settings related to machine learning.

To disable AI in Firefox, toggle the browser.ml.enabled setting to false. This should disable all AI features across the board, but community forums suggest this isn’t always enough to do the trick. For a scorched-earth approach, set the following parameters to false (or selectively keep only what you need):

• ml.chat.enabled
• ml.linkPreview.enabled
• ml.pageAssist.enabled
• ml.smartAssist.enabled
• ml.enabled
• ai.control.translations
• tabs.groups.smart.enabled
• urlbar.quicksuggest.mlEnabled

This will kill off chatbot integrations, AI-generated link descriptions, assistants and extensions, local translation of websites, tab grouping, and other AI-driven features.

How to disable AI features in Microsoft apps

Microsoft has managed to bake AI into almost every single one of its products, and turning it off is often no easy task — especially since the AI sometimes has a habit of resurrecting itself without your involvement.

How to disable AI features in Edge

Microsoft’s browser is packed with AI features, ranging from Copilot to automated search. To shut them down, follow the same logic as with Chrome: type edge://flags into the Edge address bar, hit Enter, then type “AI” or “Copilot” into the search box. From there, you can toggle off the unwanted AI features, such as:

• Enable Compose (AI-writing) on the web
• Edge Copilot Mode
• Edge History AI

Another way to ditch Copilot is to enter edge://settings/appearance/copilotAndSidebar into the address bar. Here, you can customize the look of the Copilot sidebar and tweak personalization options for results and notifications. Don’t forget to peek into the Copilot section under App-specific settings — you’ll find some additional controls tucked away there.

How to disable Microsoft Copilot

Microsoft Copilot comes in two flavors: as a component of Windows (Microsoft Copilot), and as part of the Office suite (Microsoft 365 Copilot). Their functions are similar, but you’ll have to disable one or both depending on exactly what the Redmond engineers decided to shove onto your machine.

The simplest thing you can do is just uninstall the app entirely. Right-click the Copilot entry in the Start menu and select Uninstall. If that option isn’t there, head over to your installed apps list (Start → Settings → Apps) and uninstall Copilot from there.

In certain builds of Windows 11, Copilot is baked directly into the OS, so a simple uninstall might not work. In that case, you can toggle it off via the settings: Start → Settings → Personalization → Taskbar → turn off Copilot.

If you ever have a change of heart, you can always reinstall Copilot from the Microsoft Store.

It’s worth noting that many users have complained about Copilot automatically reinstalling itself, so you might want to do a weekly check for a couple of months to make sure it hasn’t staged a comeback. For those who are comfortable tinkering with the System Registry (and understand the consequences), you can follow this detailed guide to prevent Copilot’s silent resurrection by disabling the SilentInstalledAppsEnabled flag and adding/enabling the TurnOffWindowsCopilot parameter.

How to disable Microsoft Recall

The Microsoft Recall feature, first introduced in 2024, works by constantly taking screenshots of your computer screen and having a neural network analyze them. All that extracted information is dumped into a database, which you can then search using an AI assistant. We’ve previously written in detail about the massive security risks Microsoft Recall poses.

Under pressure from cybersecurity experts, Microsoft was forced to push the launch of this feature from 2024 to 2025, significantly beefing up the protection of the stored data. However, the core of Recall remains the same: your computer still remembers your every move by constantly snapping screenshots and OCR-ing the content. And while the feature is no longer enabled by default, it’s absolutely worth checking to make sure it hasn’t been activated on your machine.

To check, head to the settings: Start → Settings → Privacy & Security → Recall & snapshots. Ensure the Save snapshots toggle is turned off, and click Delete snapshots to wipe any previously collected data, just in case.

You can also check out our detailed guide on how to disable and completely remove Microsoft Recall.

How to disable AI in Notepad and Windows context actions

AI has seeped into every corner of Windows, even into File Explorer and Notepad. You might even trigger AI features just by accidentally highlighting text in an app — a feature Microsoft calls “AI Actions”. To shut this down, head to Start → Settings → Privacy & Security → Click to Do.

Notepad has received its own special Copilot treatment, so you’ll need to disable AI there separately. Open the Notepad settings, find the AI features section, and toggle Copilot off.

Finally, Microsoft has even managed to bake Copilot into Paint. Unfortunately, as of right now, there is no official way to disable the AI features within the Paint app itself.

How to disable AI in WhatsApp

In several regions, WhatsApp users have started seeing typical AI additions like suggested replies, AI message summaries, and a brand-new Chat with Meta AI button. While Meta claims the first two features process data locally on your device and don’t ship your chats off to their servers, verifying that is no small feat. Luckily, turning them off is straightforward.

To disable Suggested Replies, go to Settings → Chats → Suggestions & smart replies and toggle off Suggested replies. You can also kill off AI Sticker suggestions in that same menu. As for the AI message summaries, those are managed in a different location: Settings → Notifications → AI message summaries.

How to disable AI on Android

Given the sheer variety of manufacturers and Android flavors, there’s no one-size-fits-all instruction manual for every single phone. Today, we’ll focus on killing off Google’s AI services — but if you’re using a device from Samsung, Xiaomi, or others, don’t forget to check your specific manufacturer’s AI settings. Just a heads-up: fully scrubbing every trace of AI might be a tall order — if it’s even possible at all.

In Google Messages, the AI features are tucked away in the settings: tap your account picture, select Messages settings, then Gemini in Messages, and toggle the assistant off.

Broadly speaking, the Gemini chatbot is a standalone app that you can uninstall by heading to your phone’s settings and selecting Apps. However, given Google’s master plan to replace the long-standing Google Assistant with Gemini, uninstalling it might become difficult — or even impossible — down the road.

If you can’t completely uninstall Gemini, head into the app to kill its features manually. Tap your profile icon, select Gemini Apps activity, and then choose Turn off or Turn off and delete activity. Next, tap the profile icon again and go to the Connected Apps setting (it may be hiding under the Personal Intelligence setting). From here, you should disable all the apps where you don’t want Gemini poking its nose in.

• For more on dealing with preloaded and system apps, check out our post, Delete the undeletable: how to disable and remove Android bloatware.

How to disable AI in macOS and iOS

Apple’s platform-level AI features, collectively known as Apple Intelligence, are refreshingly straightforward to disable. In your settings — on desktops, smartphones, and tablets alike — simply look for the section labeled Apple Intelligence & Siri. By the way, depending on your region and the language you’ve selected for your OS and Siri, Apple Intelligence might not even be available to you yet.

Other posts to help you tune the AI tools on your devices:

• Privacy settings in ChatGPT
• DeepSeek: configuring privacy and deploying a local version
• The pros and cons of AI-powered browsers
• Is a Gemini AI update about to kill privacy on your Android device?
• Should you disable Microsoft Recall in 2025?

Everyone has likely heard of OpenClaw, previously known as “Clawdbot” or “Moltbot”, the open-source AI assistant that can be deployed on a machine locally. It plugs into popular chat platforms like WhatsApp, Telegram, Signal, Discord, and Slack, which allows it to accept commands from its owner and go to town on the local file system. It has access to the owner’s calendar, email, and browser, and can even execute OS commands via the shell.

From a security perspective, that description alone should be enough to give anyone a nervous twitch. But when people start trying to use it for work within a corporate environment, anxiety quickly hardens into the conviction of imminent chaos. Some experts have already dubbed OpenClaw the biggest insider threat of 2026. The issues with OpenClaw cover the full spectrum of risks highlighted in the recent OWASP Top 10 for Agentic Applications.

OpenClaw permits plugging in any local or cloud-based LLM, and the use of a wide range of integrations with additional services. At its core is a gateway that accepts commands via chat apps or a web UI, and routes them to the appropriate AI agents. The first iteration, dubbed Clawdbot, dropped in November 2025; by January 2026, it had gone viral — and brought a heap of security headaches with it. In a single week, several critical vulnerabilities were disclosed, malicious skills cropped up in the skill directory, and secrets were leaked from Moltbook (essentially “Reddit for bots”). To top it off, Anthropic issued a trademark demand to rename the project to avoid infringing on “Claude”, and the project’s X account name was hijacked to shill crypto scams.

Known OpenClaw issues

Though the project’s developer appears to acknowledge that security is important, since this is a hobbyist project there are zero dedicated resources for vulnerability management or other product security essentials.

OpenClaw vulnerabilities

Among the known vulnerabilities in OpenClaw, the most dangerous is CVE-2026-25253 (CVSS 8.8). Exploiting it leads to a total compromise of the gateway, allowing an attacker to run arbitrary commands. To make matters worse, it’s alarmingly easy to pull off: if the agent visits an attacker’s site or the user clicks a malicious link, the primary authentication token is leaked. With that token in hand, the attacker has full administrative control over the gateway. This vulnerability was patched in version 2026.1.29.

Also, two dangerous command injection vulnerabilities (CVE-2026-24763 and CVE-2026-25157) were discovered.

Insecure defaults and features

A variety of default settings and implementation quirks make attacking the gateway a walk in the park:

• Authentication is disabled by default, so the gateway is accessible from the internet.
• The server accepts WebSocket connections without verifying their origin.
• Localhost connections are implicitly trusted, which is a disaster waiting to happen if the host is running a reverse proxy.
• Several tools — including some dangerous ones — are accessible in Guest Mode.
• Critical configuration parameters leak across the local network via mDNS broadcast messages.

Secrets in plaintext

OpenClaw’s configuration, “memory”, and chat logs store API keys, passwords, and other credentials for LLMs and integration services in plain text. This is a critical threat — to the extent that versions of the RedLine and Lumma infostealers have already been spotted with OpenClaw file paths added to their must-steal lists. Also, the Vidar infostealer was caught stealing secrets from OpenClaw.

Malicious skills

OpenClaw’s functionality can be extended with “skills” available in the ClawHub repository. Since anyone can upload a skill, it didn’t take long for threat actors to start “bundling” the AMOS macOS infostealer into their uploads. Within a short time, the number of malicious skills reached the hundreds. This prompted developers to quickly ink a deal with VirusTotal to ensure all uploaded skills aren’t only checked against malware databases, but also undergo code and content analysis via LLMs. That said, the authors are very clear: it’s no silver bullet.

Structural flaws in the OpenClaw AI agent

Vulnerabilities can be patched and settings can be hardened, but some of OpenClaw’s issues are fundamental to its design. The product combines several critical features that, when bundled together, are downright dangerous:

• OpenClaw has privileged access to sensitive data on the host machine and the owner’s personal accounts.
• The assistant is wide open to untrusted data: the agent receives messages via chat apps and email, autonomously browses web pages, etc.
• It suffers from the inherent inability of LLMs to reliably separate commands from data, making prompt injection a possibility.
• The agent saves key takeaways and artifacts from its tasks to inform future actions. This means a single successful injection can poison the agent’s memory, influencing its behavior long-term.
• OpenClaw has the power to talk to the outside world — sending emails, making API calls, and utilizing other methods to exfiltrate internal data.

It’s worth noting that while OpenClaw is a particularly extreme example, this “Terrifying Five” list is actually characteristic of almost all multi-purpose AI agents.

OpenClaw risks for organizations

If an employee installs an agent like this on a corporate device and hooks it into even a basic suite of services (think Slack and SharePoint), the combination of autonomous command execution, broad file system access, and excessive OAuth permissions creates fertile ground for a deep network compromise. In fact, the bot’s habit of hoarding unencrypted secrets and tokens in one place is a disaster waiting to happen — even if the AI agent itself is never compromised.

On top of that, these configurations violate regulatory requirements across multiple countries and industries, leading to potential fines and audit failures. Current regulatory requirements, like those in the EU AI Act or the NIST AI Risk Management Framework, explicitly mandate strict access control for AI agents. OpenClaw’s configuration approach clearly falls short of those standards.

But the real kicker is that even if employees are banned from installing this software on work machines, OpenClaw can still end up on their personal devices. This also creates specific risks for given the organization as a whole:

• Personal devices frequently store access to work systems like corporate VPN configs or browser tokens for email and internal tools. These can be hijacked to gain a foothold in the company’s infrastructure.
• Controlling the agent via chat apps means that it’s not just the employee that becomes a target for social engineering, but also their AI agent, seeing AI account takeovers or impersonation of the user in chats with colleagues (among other scams) become a reality. Even if work is only occasionally discussed in personal chats, the info in them is ripe for the picking.
• If an AI agent on a personal device is hooked into any corporate services (email, messaging, file storage), attackers can manipulate the agent to siphon off data, and this activity would be extremely difficult for corporate monitoring systems to spot.

How to detect OpenClaw

Depending on the SOC team’s monitoring and response capabilities, they can track OpenClaw gateway connection attempts on personal devices or in the cloud. Additionally, a specific combination of red flags can indicate OpenClaw’s presence on a corporate device:

• Look for ~/.openclaw/, ~/clawd/, or ~/.clawdbot directories on host machines.
• Scan the network with internal tools, or public ones like Shodan, to identify the HTML fingerprints of Clawdbot control panels.
• Monitor for WebSocket traffic on ports 3000 and 18789.
• Keep an eye out for mDNS broadcast messages on port 5353 (specifically openclaw-gw.tcp).
• Watch for unusual authentication attempts in corporate services, such as new App ID registrations, OAuth Consent events, or User-Agent strings typical of Node.js and other non-standard user agents.
• Look for access patterns typical of automated data harvesting: reading massive chunks of data (scraping all files or all emails) or scanning directories at fixed intervals during off-hours.

Controlling shadow AI

A set of security hygiene practices can effectively shrink the footprint of both shadow IT and shadow AI, making it much harder to deploy OpenClaw in an organization:

• Use host-level allowlisting to ensure only approved applications and cloud integrations are installed. For products that support extensibility (like Chrome extensions, VS Code plugins, or OpenClaw skills), implement a closed list of vetted add-ons.
• Conduct a full security assessment of any product or service, AI agents included, before allowing them to hook into corporate resources.
• Treat AI agents with the same rigorous security requirements applied to public-facing servers that process sensitive corporate data.
• Implement the principle of least privilege for all users and other identities.
• Don’t grant administrative privileges without a critical business need. Require all users with elevated permissions to use them only when performing specific tasks rather than working from privileged accounts all the time.
• Configure corporate services so that technical integrations (like apps requesting OAuth access) are granted only the bare minimum permissions.
• Periodically audit integrations, OAuth tokens, and permissions granted to third-party apps. Review the need for these with business owners, proactively revoke excessive permissions, and kill off stale integrations.

Secure deployment of agentic AI

If an organization allows AI agents in an experimental capacity — say, for development testing or efficiency pilots — or if specific AI use cases have been greenlit for general staff, robust monitoring, logging, and access control measures should be implemented:

• Deploy agents in an isolated subnet with strict ingress and egress rules, limiting communication only to trusted hosts required for the task.
• Use short-lived access tokens with a strictly limited scope of privileges. Never hand an agent tokens that grant access to core company servers or services. Ideally, create dedicated service accounts for every individual test.
• Wall off the agent from dangerous tools and data sets that aren’t relevant to its specific job. For experimental rollouts, it’s best practice to test the agent using purely synthetic data that mimics the structure of real production data.
• Configure detailed logging of the agent’s actions. This should include event logs, command-line parameters, and chain-of-thought artifacts associated with every command it executes.
• Set up SIEM to flag abnormal agent activity. The same techniques and rules used to detect LotL attacks are applicable here, though additional efforts to define what normal activity looks like for a specific agent are required.
• If MCP servers and additional agent skills are used, scan them with the security tools emerging for these tasks, such as skill-scanner, mcp-scanner, or mcp-scan. Specifically for OpenClaw testing, several companies have already released open-source tools to audit the security of its configurations.

Corporate policies and employee training

A flat-out ban on all AI tools is a simple but rarely productive path. Employees usually find workarounds — driving the problem into the shadows where it’s even harder to control. Instead, it’s better to find a sensible balance between productivity and security.

Implement transparent policies on using agentic AI. Define which data categories are okay for external AI services to process, and which are strictly off-limits. Employees need to understand why something is forbidden. A policy of “yes, but with guardrails” is always received better than a blanket “no”.

Train with real-world examples. Abstract warnings about “leakage risks” tend to be futile. It’s better to demonstrate how an agent with email access can forward confidential messages just because a random incoming email asked it to. When the threat feels real, motivation to follow the rules grows too. Ideally, employees should complete a brief crash course on AI security.

Offer secure alternatives. If employees need an AI assistant, provide an approved tool that features centralized management, logging, and OAuth access control.

To implement effective cybersecurity programs and keep the security team deeply integrated into all business processes, the CISO needs to regularly demonstrate the value of this work to senior management. This requires speaking the language of business, but a dangerous trap awaits those who try.  Security professionals and executives often use the same words, but for entirely different things. Sometimes, a number of similar terms are used interchangeably. As a result, top management may not understand which threats the security team is trying to mitigate, what the company’s actual level of cyber-resilience is, or where budget and resources are being allocated. Therefore, before presenting sleek dashboards or calculating the ROI of security programs, it’s worth subtly clarifying these important terminological nuances.

By clarifying these terms and building a shared vocabulary, the CISO and the Board can significantly improve communication and, ultimately, strengthen the organization’s overall security posture.

Why cybersecurity vocabulary matters for management

Varying interpretations of terms are more than just an inconvenience; the consequences can be quite substantial. A lack of clarity regarding details can lead to:

• Misallocated investments. Management might approve the purchase of a zero trust solution without realizing it’s only one piece of a long-term, comprehensive program with a significantly larger budget. The money is spent, yet the results management expected are never achieved. Similarly, with regard to cloud migration, management may assume that moving to the cloud automatically transfers all security responsibility to the provider, and subsequently reject the cloud security budget.
• Blind acceptance of risk. Business unit leaders may accept cybersecurity risks without having a full understanding of the potential impact.
• Lack of governance. Without understanding the terminology, management can’t ask the right — tough — questions, or assign areas of responsibility effectively. When an incident occurs, it often turns out that business owners believed security was entirely within the CISO’s domain, while the CISO lacked the authority to influence business processes.

Cyber-risk vs. IT risk

Many executives believe that cybersecurity is a purely technical issue they can hand off to IT. Even though the importance of cybersecurity to business is indisputable, and cyber-incidents have long ranked as a top business risk, surveys show that many organizations still fail to engage non-technical leaders in cybersecurity discussions.

Information security risks are often lumped in with IT concerns like uptime and service availability.  In reality, cyberrisk is a strategic business risk linked to business continuity, financial loss, and reputational damage.

IT risks are generally operational in nature, affecting efficiency, reliability, and cost management. Responding to IT incidents is often handled entirely by IT staff. Major cybersecurity incidents, however, have a much broader scope; they require the engagement of nearly every department, and have a long-term impact on the organization in many ways — including as regards reputation, regulatory compliance, customer relationships, and overall financial health.

Compliance vs. security

Cybersecurity is integrated into regulatory requirements at every level — from international directives like NIS2 and GDPR, to cross-border industry guidelines like PCI DSS, plus specific departmental mandates. As a result, company management often views cybersecurity measures as compliance checkboxes, believing that once regulatory requirements are met, cybersecurity issues can be considered resolved. This mindset can stem from a conscious effort to minimize security spending (“we’re not doing more than what we’re required to”) or from a sincere misunderstanding (“we’ve passed an ISO 27001 audit, so we’re unhackable”).

In reality, compliance is meeting the minimum requirements of auditors and government regulators at a specific point in time. Unfortunately, the history of large-scale cyberattacks on major organizations proves that “minimum” requirements have that name for a reason. For real protection against modern cyberthreats, companies must continuously improve their security strategies and measures according to the specific needs of the given industry.

Threat, vulnerability, and risk

These three terms are often used synonymously, which leads to erroneous conclusions made by management: “There’s a critical vulnerability on our server? That means we have a critical risk!” To avoid panic or, conversely, inaction, it’s vital to use these terms precisely and understand how they relate to one another.

A vulnerability is a weakness — an “open door”. This could be a flaw in software code, a misconfigured server, an unlocked server room, or an employee who opens every email attachment.

A threat is a potential cause of an incident. This could be a malicious actor, malware, or even a natural disaster. A threat is what might “walk through that open door”.

Risk is the potential loss. It’s the cumulative assessment of the likelihood of a successful attack, and what the organization stands to lose as a result (the impact).

The connections among these elements are best explained with a simple formula:

Risk = (Threat × Vulnerability) × Impact

This can be illustrated as follows. Imagine a critical vulnerability with a maximum severity rating is discovered in an outdated system. However, this system is disconnected from all networks, sits in an isolated room, and is handled by only three vetted employees. The probability of an attacker reaching it is near zero. Meanwhile, the lack of two-factor authentication in the accounting systems creates a real, high risk, resulting from both a high probability of attack and significant potential damage.

Incident response, disaster recovery, and business continuity

Management’s perception of security crises is often oversimplified: “If we get hit by ransomware, we’ll just activate the IT Disaster Recovery plan and restore from backups”. However, conflating these concepts — and processes — is extremely dangerous.

Incident Response (IR) is the responsibility of the security team or specialist contractors. Their job is to localize the threat, kick the attacker out of the network, and stop the attack from spreading.

Disaster Recovery (DR) is an IT engineering task. It’s the process of restoring servers and data from backups after the incident response has been completed.

Business Continuity (BC) is a strategic task for top management. It’s the plan for how the company continues to serve customers, ship goods, pay compensation, and talk to the press while its primary systems are still offline.

If management focuses solely on recovery, the company will lack an action plan for the most critical period of downtime.

Security awareness vs. security culture

Leaders at all levels sometimes assume that simply conducting security training guarantees results: “The employees have passed their annual test, so now they won’t click on a phishing link”. Unfortunately, relying solely on training organized by HR and IT won’t cut it. Effectiveness requires changing the team’s behavior, which is impossible without the engagement of business management.

Awareness is knowledge. An employee knows what phishing is and understands the importance of complex passwords.

Security culture refers to behavioral patterns. It’s what an employee does in a stressful situation or when no one’s watching. Culture isn’t shaped by tests, but by an environment where it’s safe to report mistakes and where it’s customary to identify and prevent potentially dangerous situations. If an employee fears punishment, they’ll hide an incident. In a healthy culture, they’ll report a suspicious email to the SOC, or nudge a colleague who forgets to lock their computer, thereby becoming an active link in the defense chain.

Detection vs. prevention

Business leaders often think in outdated “fortress wall” categories: “We bought expensive protection systems, so there should be no way to hack us. If an incident occurs, it means the CISO failed”. In practice, preventing 100% of attacks is technically impossible and economically prohibitive. Modern strategy is built on a balance between cybersecurity and business effectiveness. In a balanced system, components focused on threat detection and prevention work in tandem.

Prevention deflects automated, mass attacks.

Detection and Response help identify and neutralize more professional, targeted attacks that manage to bypass prevention tools or exploit vulnerabilities.

The key objective of the cybersecurity team today isn’t to guarantee total invulnerability, but to detect an attack at an early stage and minimize the impact on the business. To measure success here, the industry typically uses metrics like Mean Time to Detect (MTTD) and Mean Time to Respond (MTTR).

Zero-trust philosophy vs. zero-trust products

The zero trust concept — which implies “never trust, always verify” for all components of IT infrastructure — has long been recognized as relevant and effective in corporate security. It requires constant verification of identity (user accounts, devices, and services) and context for every access request based on the assumption that the network has already been compromised.

However, the presence of “zero trust” in the name of a security solution doesn’t mean an organization can adopt this approach overnight simply by purchasing the product.
Zero trust isn’t a product you can “turn on”; it’s an architectural strategy and a long-term transformation journey. Implementing zero trust requires restructuring access processes and refining IT systems to ensure continuous verification of identity and devices. Buying software without changing processes won’t have a significant effect.

Security of the cloud vs. security in the cloud

When migrating IT services to cloud infrastructure like AWS or Azure, there’s often an illusion of a total risk transfer: “We pay the provider, so security is now their headache”. This is a dangerous misconception, and a misinterpretation of what is known as the Shared Responsibility Model.

Security of the cloud is the provider’s responsibility. It protects the data centers, the physical servers, and the cabling.

Security in the cloud is the client’s responsibility.

Discussions regarding budgets for cloud projects and their security aspects should be accompanied by real life examples. The provider protects the database from unauthorized access according to the settings configured by the client’s employees. If employees leave a database open or use weak passwords, and if two-factor authentication isn’t enabled for the administrator panel, the provider can’t prevent unauthorized individuals from downloading the information — an all-too-common news story. Therefore, the budget for these projects must account for cloud security tools and configuration management on the company side.

Vulnerability scanning vs. penetration testing

Leaders often confuse automated checks, which fall under cyber-hygiene, with assessing IT assets for resilience against sophisticated attacks: “Why pay hackers for a pentest when we run the scanner every week?”

Vulnerability scanning checks a specific list of IT assets for known vulnerabilities. To put it simply, it’s like a security guard doing the rounds to check that the office windows and doors are locked.

Penetration testing (pentesting) is a manual assessment to evaluate the possibility of a real-world breach by exploiting vulnerabilities. To continue the analogy, it’s like hiring an expert burglar to actually try and break into the office.

One doesn’t replace the other; to understand its true security posture, a business needs both tools.

Managed assets vs. attack surface

A common and dangerous misconception concerns the scope of protection and the overall visibility held by IT and Security. A common refrain at meetings is, “We have an accurate inventory list of our hardware. We’re protecting everything we own”.

Managed IT assets are things the IT department has purchased, configured, and can see in their reports.

An attack surface is anything accessible to attackers: any potential entry point into the company. This includes Shadow IT (cloud services, personal messaging apps, test servers…), which is basically anything employees launch themselves in circumvention of official protocols to speed up or simplify their work. Often, it’s these “invisible” assets that become the entry point for an attack, as the security team can’t protect what it doesn’t know exists.

The year 2025 saw a record-breaking number of attacks on Android devices. Scammers are currently riding a few major waves: the hype surrounding AI apps, the urge to bypass site blocks or age checks, the hunt for a bargain on a new smartphone, the ubiquity of mobile banking, and, of course, the popularity of NFC. Let’s break down the primary threats of 2025–2026, and figure out how to keep your Android device safe in this new landscape.

Sideloading

Malicious installation packages (APK files) have always been the Final Boss among Android threats, despite Google’s multi-year efforts to fortify the OS. By using sideloading — installing an app via an APK file instead of grabbing it from the official store — users can install pretty much anything, including straight-up malware. And neither the rollout of Google Play Protect, nor the various permission restrictions for shady apps have managed to put a dent in the scale of the problem.

According to preliminary data from Kaspersky for 2025, the number of detected Android threats grew almost by half. In the third quarter alone, detections jumped by 38% compared to the second. In certain niches, like Trojan bankers, the growth was even more aggressive. In Russia alone, the notorious Mamont banker attacked 36 times more users than it did the previous year, while globally this entire category saw a nearly fourfold increase.

Today, bad actors primarily distribute malware via messaging apps by sliding malicious files into DMs and group chats. The installation file usually sports an enticing name (think “party_pics.jpg.apk” or “clearance_sale_catalog.apk”), accompanied by a message “helpfully” explaining how to install the package while bypassing the OS restrictions and security warnings.

Once a new device is infected, the malware often spams itself to everyone in the victim’s contact list.

Search engine spam and email campaigns are also trending, luring users to sites that look exactly like an official app store. There, they’re prompted to download the “latest helpful app”, such as an AI assistant. In reality, instead of an installation from an official app store, the user ends up downloading an APK package. A prime example of these tactics is the ClayRat Android Trojan, which uses a mix of all these techniques to target Russian users. It spreads through groups and fake websites, blasts itself to the victim’s contacts via SMS, and then proceeds to steal the victim’s chat logs and call history; it even goes as far as snapping photos of the owner using the front-facing camera. In just three months, over 600 distinct ClayRat builds have surfaced.

The scale of the disaster is so massive that Google even announced an upcoming ban on distributing apps from unknown developers starting in 2026. However, after a couple of months of pushback from the dev community, the company pivoted to a softer approach: unsigned apps will likely only be installable via some kind of superuser mode. As a result, we can expect scammers to simply update their how-to guides with instructions on how to toggle that mode on.

Kaspersky for Android will help you protect yourself from counterfeit and trojanized APK files. Unfortunately, due to Google’s decision, our Android security apps are currently unavailable on Google Play. We’ve previously provided detailed information on how to install our Android apps with a 100% guarantee of authenticity.

NFC relay attacks

Once an Android device is compromised, hackers can skip the middleman to steal the victim’s money directly thanks to the massive popularity of mobile payments. In the third quarter of 2025 alone, over 44 000 of these attacks were detected in Russia alone — a 50% jump from the previous quarter.

There are two main scams currently in play: direct and reverse NFC exploits.

Direct NFC relay is when a scammer contacts the victim via a messaging app and convinces them to download an app — supposedly to “verify their identity” with their bank. If the victim bites and installs it, they’re asked to tap their physical bank card against the back of their phone and enter their PIN. And just like that the card data is handed over to the criminals, who can then drain the account or go on a shopping spree.

Reverse NFC relay is a more elaborate scheme. The scammer sends a malicious APK and convinces the victim to set this new app as their primary contactless payment method. The app generates an NFC signal that ATMs recognize as the scammer’s card. The victim is then talked into going to an ATM with their infected phone to deposit cash into a “secure account”. In reality, those funds go straight into the scammer’s pocket.

We break both of these methods down in detail in our post, NFC skimming attacks.

NFC is also being leveraged to cash out cards after their details have been siphoned off through phishing websites. In this scenario, attackers attempt to link the stolen card to a mobile wallet on their own smartphone — a scheme we covered extensively in NFC carders hide behind Apple Pay and Google Wallet.

The stir over VPNs

In many parts of the world, getting onto certain websites isn’t as simple as it used to be. Some sites are blocked by local internet regulators or ISPs via court orders; others require users to pass an age verification check by showing ID and personal info. In some cases, sites block users from specific countries entirely just to avoid the headache of complying with local laws. Users are constantly trying to bypass these restrictions —and they often end up paying for it with their data or cash.

Many popular tools for bypassing blocks — especially free ones — effectively spy on their users. A recent audit revealed that over 20 popular services with a combined total of more than 700 million downloads actively track user location. They also tend to use sketchy encryption at best, which essentially leaves all user data out in the open for third parties to intercept.

Moreover, according to Google data from November 2025, there was a sharp spike in cases where malicious apps are being disguised as legitimate VPN services to trick unsuspecting users.

The permissions that this category of apps actually requires are a perfect match for intercepting data and manipulating website traffic. It’s also much easier for scammers to convince a victim to grant administrative privileges to an app responsible for internet access than it is for, say, a game or a music player. We should expect this scheme to only grow in popularity.

Trojan in a box

Even cautious users can fall victim to an infection if they succumb to the urge to save some cash. Throughout 2025, cases were reported worldwide where devices were already carrying a Trojan the moment they were unboxed. Typically, these were either smartphones from obscure manufacturers or knock-offs of famous brands purchased on online marketplaces. But the threat wasn’t limited to just phones; TV boxes, tablets, smart TVs, and even digital photo frames were all found to be at risk.

It’s still not entirely clear whether the infection happens right on the factory floor or somewhere along the supply chain between the factory and the buyer’s doorstep, but the device is already infected before the first time it’s turned on. Usually, it’s a sophisticated piece of malware called Triada, first identified by Kaspersky analysts back in 2016. It’s capable of injecting itself into every running app to intercept information: stealing access tokens and passwords for popular messaging apps and social media, hijacking SMS messages (confirmation codes: ouch!), redirecting users to ad-heavy sites, and even running a proxy directly on the phone so attackers can browse the web using the victim’s identity.

Technically, the Trojan is embedded right into the smartphone’s firmware, and the only way to kill it is to reflash the device with a clean OS. Usually, once you dig into the system, you’ll find that the device has far less RAM or storage than advertised — meaning the firmware is literally lying to the owner to sell a cheap hardware config as something more premium.

Another common pre-installed menace is the BADBOX 2.0 botnet, which also pulls double duty as a proxy and an ad-fraud engine. This one specializes in TV boxes and similar hardware.

How to go on using Android without losing your mind

Despite the growing list of threats, you can still use your Android smartphone safely! You just have to stick to some strict mobile hygiene rules.

• Install a comprehensive security solution on all your smartphones. We recommend Kaspersky for Android to protect against malware and phishing.
• Avoid sideloading apps via APKs whenever you can use an app store instead. A known app store — even a smaller one — is always a better bet than a random APK from some random website. If you have no other choice, download APK files only from official company websites, and double-check the URL of the page you’re on. If you aren’t 100% sure what the official site is, don’t just rely on a search engine; check official business directories or at least Wikipedia to verify the correct address.
• Read OS warnings carefully during installation. Don’t grant permissions if the requested rights or actions seem illogical or excessive for the app you’re installing.
• Under no circumstances should you install apps from links or attachments in chats, emails, or similar communication channels.
• Never tap your physical bank card against your phone. There is absolutely no legitimate scenario where doing this would be for your own benefit.
• Do not enter your card’s PIN into any app on your phone. A PIN should only ever be requested by an ATM or a physical payment terminal.
• When choosing a VPN, stick to paid ones from reputable companies.
• Buy smartphones and other electronics from official retailers, and steer clear of brands you’ve never heard of. Remember: if a deal seems too good to be true, it almost certainly is.

Other major Android threats from 2025:

• Pixnapping vulnerability: unblockable screenshots of your Android phone
• Spyware that pretends to be an antivirus
• Data theft during smartphone charging
• SparkCat trojan stealer infiltrates App Store and Google Play, steals data from photos

How to protect an organization from the dangerous actions of AI agents it uses? This isn’t just a theoretical what-if anymore — considering the actual damage autonomous AI can do ranges from providing poor customer service to destroying corporate primary databases.  It’s a question business leaders are currently hammering away at, and government agencies and security experts are racing to provide answers to.

For CIOs and CISOs, AI agents create a massive governance headache. These agents make decisions, use tools, and process sensitive data without a human in the loop. Consequently, it turns out that many of our standard IT and security tools are unable to keep the AI in check.

The non-profit OWASP Foundation has released a handy playbook on this very topic. Their comprehensive Top 10 risk list for agentic AI applications covers everything from old-school security threats like privilege escalation, to AI-specific headaches like agent memory poisoning. Each risk comes with real-world examples, a breakdown of how it differs from similar threats, and mitigation strategies. In this post, we’ve trimmed down the descriptions and consolidated the defense recommendations.

Agent goal hijack (ASI01)

This risk involves manipulating an agent’s tasks or decision-making logic by exploiting the underlying model’s inability to tell the difference between legitimate instructions and external data. Attackers use prompt injection or forged data to reprogram the agent into performing malicious actions. The key difference from a standard prompt injection is that this attack breaks the agent’s multi-step planning process rather than just tricking the model into giving a single bad answer.

Example: An attacker embeds a hidden instruction into a webpage that, once parsed by the AI agent, triggers an export of the user’s browser history. A vulnerability of this very nature was showcased in a EchoLeak study.

Tool misuse and exploitation (ASI02)

This risk crops up when an agent — driven by ambiguous commands or malicious influence — uses the legitimate tools it has access to in unsafe or unintended ways. Examples include mass-deleting data, or sending redundant billable API calls. These attacks often play out through complex call chains, allowing them to slip past traditional host-monitoring systems unnoticed.

Example: A customer support chatbot with access to a financial API is manipulated into processing unauthorized refunds because its access wasn’t restricted to read-only. Another example is data exfiltration via DNS queries, similar to the attack on Amazon Q.

Identity and privilege abuse (ASI03)

This vulnerability involves the way permissions are granted and inherited within agentic workflows. Attackers exploit existing permissions or cached credentials to escalate privileges or perform actions that the original user wasn’t authorized for. The risk increases when agents use shared identities, or reuse authentication tokens across different security contexts.

Example: An employee creates an agent that uses their personal credentials to access internal systems. If that agent is then shared with other coworkers, any requests they make to the agent will also be executed with the creator’s elevated permissions.

Agentic Supply Chain Vulnerabilities (ASI04)

Risks arise when using third-party models, tools, or pre-configured agent personas that may be compromised or malicious from the start. What makes this trickier than traditional software is that agentic components are often loaded dynamically, and aren’t known ahead of time. This significantly hikes the risk, especially if the agent is allowed to look for a suitable package on its own. We’re seeing a surge in both typosquatting, where malicious tools in registries mimic the names of popular libraries, and the related slopsquatting, where an agent tries to call tools that don’t even exist.

Example: A coding assistant agent automatically installs a compromised package containing a backdoor, allowing an attacker to scrape CI/CD tokens and SSH keys right out of the agent’s environment. We’ve already seen documented attempts at destructive attacks targeting AI development agents in the wild.

Unexpected code execution / RCE (ASI05)

Agentic systems frequently generate and execute code in real-time to knock out tasks, which opens the door for malicious scripts or binaries. Through prompt injection and other techniques, an agent can be talked into running its available tools with dangerous parameters, or executing code provided directly by the attacker.  This can escalate into a full container or host compromise, or a sandbox escape — at which point the attack becomes invisible to standard AI monitoring tools.

Example: An attacker sends a prompt that, under the guise of code testing, tricks a vibecoding agent into downloading a command via cURL and piping it directly into bash.

Memory and context poisoning (ASI06)

Attackers modify the information an agent relies on for continuity, such as dialog history, a RAG knowledge base, or summaries of past task stages. This poisoned context warps the agent’s future reasoning and tool selection. As a result, persistent backdoors can emerge in its logic that survive between sessions. Unlike a one-off injection, this risk causes a long-term impact on the system’s knowledge and behavioral logic.

Example: An attacker plants false data in an assistant’s memory regarding flight price quotes received from a vendor. Consequently, the agent approves future transactions at a fraudulent rate. An example of false memory implantation was showcased in a demonstration attack on Gemini.

Insecure inter-agent communication (ASI07)

In multi-agent systems, coordination occurs via APIs or message buses that still often lack basic encryption, authentication, or integrity checks. Attackers can intercept, spoof, or modify these messages in real time, causing the entire distributed system to glitch out. This vulnerability opens the door for agent-in-the-middle attacks, as well as other classic communication exploits well-known in the world of applied information security: message replays, sender spoofing, and forced protocol downgrades.

Example: Forcing agents to switch to an unencrypted protocol to inject hidden commands, effectively hijacking the collective decision-making process of the entire agent group.

Cascading failures (ASI08)

This risk describes how a single error — caused by hallucination, a prompt injection, or any other glitch — can ripple through and amplify across a chain of autonomous agents. Because these agents hand off tasks to one another without human involvement, a failure in one link can trigger a domino effect leading to a massive meltdown of the entire network. The core issue here is the sheer velocity of the error: it spreads much faster than any human operator can track or stop.

Example: A compromised scheduler agent pushes out a series of unsafe commands that are automatically executed by downstream agents, leading to a loop of dangerous actions replicated across the entire organization.

Human–agent trust exploitation (ASI09)

Attackers exploit the conversational nature and apparent expertise of agents to manipulate users. Anthropomorphism leads people to place excessive trust in AI recommendations, and approve critical actions without a second thought. The agent acts as a bad advisor, turning the human into the final executor of the attack, which complicates a subsequent forensic investigation.

Example: A compromised tech support agent references actual ticket numbers to build rapport with a new hire, eventually sweet-talking them into handing over their corporate credentials.

Rogue agents (ASI10)

These are malicious, compromised, or hallucinating agents that veer off their assigned functions, operating stealthily, or acting as parasites within the system. Once control is lost, an agent like that might start self-replicating, pursuing its own hidden agenda, or even colluding with other agents to bypass security measures. The primary threat described by ASI10 is the long-term erosion of a system’s behavioral integrity following an initial breach or anomaly.

Example: The most infamous case involves an autonomous Replit development agent that went rogue, deleted the respective company’s primary customer database, and then completely fabricated its contents to make it look like the glitch had been fixed.

Mitigating risks in agentic AI systems

While the probabilistic nature of LLM generation and the lack of separation between instructions and data channels make bulletproof security impossible, a rigorous set of controls — approximating a Zero Trust strategy — can significantly limit the damage when things go awry. Here are the most critical measures.

Enforce the principles of both least autonomy and least privilege. Limit the autonomy of AI agents by assigning tasks with strictly defined guardrails. Ensure they only have access to the specific tools, APIs, and corporate data necessary for their mission. Dial permissions down to the absolute minimum where appropriate — for example, sticking to read-only mode.

Use short-lived credentials. Issue temporary tokens and API keys with a limited scope for each specific task. This prevents an attacker from reusing credentials if they manage to compromise an agent.

Mandatory human-in-the-loop for critical operations. Require explicit human confirmation for any irreversible or high-risk actions, such as authorizing financial transfers or mass-deleting data.

Execution isolation and traffic control. Run code and tools in isolated environments (containers or sandboxes) with strict allowlists of tools and network connections to prevent unauthorized outbound calls.

Policy enforcement. Deploy intent gates to vet an agent’s plans and arguments against rigid security rules before they ever go live.

Input and output validation and sanitization. Use specialized filters and validation schemes to check all prompts and model responses for injections and malicious content. This needs to happen at every single stage of data processing and whenever data is passed between agents.

Continuous secure logging. Record every agent action and inter-agent message in immutable logs. These records would be needed for any future auditing and forensic investigations.

Behavioral monitoring and watchdog agents. Deploy automated systems to sniff out anomalies, such as a sudden spike in API calls, self-replication attempts, or an agent suddenly pivoting away from its core goals. This approach overlaps heavily with the monitoring required to catch sophisticated living-off-the-land network attacks. Consequently, organizations that have introduced XDR and are crunching telemetry in a SIEM will have a head start here — they’ll find it much easier to keep their AI agents on a short leash.

Supply chain control and SBOMs (software bills of materials). Only use vetted tools and models from trusted registries. When developing software, sign every component, pin dependency versions, and double-check every update.

Static and dynamic analysis of generated code. Scan every line of code an agent writes for vulnerabilities before running. Ban the use of dangerous functions like eval() completely. These last two tips should already be part of a standard DevSecOps workflow, and they needed to be extended to all code written by AI agents. Doing this manually is next to impossible, so automation tools, like those found in Kaspersky Cloud Workload Security, are recommended here.

Securing inter-agent communications. Ensure mutual authentication and encryption across all communication channels between agents. Use digital signatures to verify message integrity.

 Kill switches. Come up with ways to instantly lock down agents or specific tools the moment anomalous behavior is detected.

Using UI for trust calibration. Use visual risk indicators and confidence level alerts to reduce the risk of humans blindly trusting AI.

User training. Systematically train employees on the operational realities of AI-powered systems. Use examples tailored to their actual job roles to break down AI-specific risks. Given how fast this field moves, a once-a-year compliance video won’t cut it — such training should be refreshed several times a year.

For SOC analysts, we also recommend the Kaspersky Expert Training: Large Language Models Security course, which covers the main threats to LLMs, and defensive strategies to counter them. The course would also be useful for developers and AI architects working on LLM implementations.

A newly discovered vulnerability named WhisperPair can turn Bluetooth headphones and headsets from many well-known brands into personal tracking beacons — regardless of whether the accessories are currently connected to an iPhone, Android smartphone, or even a laptop. Even though the technology behind this flaw was originally developed by Google for Android devices, the tracking risks are actually much higher for those using vulnerable headsets with other operating systems — like iOS, macOS, Windows, or Linux. For iPhone owners, this is especially concerning.

Connecting Bluetooth headphones to Android smartphones became a whole lot faster when Google rolled out Fast Pair, a technology now used by dozens of accessory manufacturers. To pair a new headset, you just turn it on and hold it near your phone. If your device is relatively modern (produced after 2019), a pop-up appears inviting you to connect and download the accompanying app, if it exists. One tap, and you’re good to go.

Unfortunately, it seems quite a few manufacturers didn’t pay attention to the particulars of this tech when implementing it, and now their accessories can be hijacked by a stranger’s smartphone in seconds — even if the headset isn’t actually in pairing mode. This is the core of the WhisperPair vulnerability, recently discovered by researchers at KU Leuven and recorded as CVE-2025-36911.

The attacking device — which can be a standard smartphone, tablet or laptop — broadcasts Google Fast Pair requests to any Bluetooth devices within a 14-meter radius. As it turns out, a long list of headphones from Sony, JBL, Redmi, Anker, Marshall, Jabra, OnePlus, and even Google itself (the Pixel Buds 2) will respond to these pings even when they aren’t looking to pair. On average, the attack takes just 10 seconds.

Once the headphones are paired, the attacker can do pretty much anything the owner can: listen in through the microphone, blast music, or — in some cases — locate the headset on a map if it supports Google Find Hub. That latter feature, designed strictly for finding lost headphones, creates a perfect opening for stealthy remote tracking. And here’s the twist: it’s actually most dangerous for Apple users and anyone else rocking non-Android hardware.

Remote tracking and the risks for iPhones

When headphones or a headset first shake hands with an Android device via the Fast Pair protocol, an owner key tied to that smartphone’s Google account is tucked away in the accessory’s memory. This info allows the headphones to be found later by leveraging data collected from millions of Android devices. If any random smartphone spots the target device nearby via Bluetooth, it reports its location to the Google servers. This feature — Google Find Hub — is essentially the Android version of Apple’s Find My, and it introduces the same unauthorized tracking risks as a rogue AirTag.

When an attacker hijacks the pairing, their key can be saved as the headset owner’s key — but only if the headset targeted via WhisperPair hasn’t previously been linked to an Android device and has only been used with an iPhone, or other hardware like a laptop with a different OS. Once the headphones are paired, the attacker can stalk their location on a map at their leisure — crucially, anywhere at all (not just within the 14-meter range).

Android users who’ve already used Fast Pair to link their vulnerable headsets are safe from this specific move, since they’re already logged in as the official owners. Everyone else, however, should probably double-check their manufacturer’s documentation to see if they’re in the clear — thankfully, not every device vulnerable to the exploit actually supports Google Find Hub.

How to neutralize the WhisperPair threat

The only truly effective way to fix this bug is to update your headphones’ firmware, provided an update is actually available. You can typically check for and install updates through the headset’s official companion app. The researchers have compiled a list of vulnerable devices on their site, but it’s almost certainly not exhaustive.

After updating the firmware, you absolutely must perform a factory reset to wipe the list of paired devices — including any unwanted guests.

If no firmware update is available and you’re using your headset with iOS, macOS, Windows, or Linux, your only remaining option is to track down an Android smartphone (or find a trusted friend who has one) and use it to reserve the role of the original owner. This will prevent anyone else from adding your headphones to Google Find Hub behind your back.

The update from Google

In January 2026, Google pushed an Android update to patch the vulnerability on the OS side. Unfortunately, the specifics haven’t been made public, so we’re left guessing exactly what they tweaked under the hood. Most likely, updated smartphones will no longer report the location of accessories hijacked via WhisperPair to the Google Find Hub network. But given that not everyone is exactly speedy when it comes to installing Android updates, it’s a safe bet that this type of headset tracking will remain viable for at least another couple of years.

Want to find out how else your gadgets might be spying on you? Check out these posts:

• How to protect yourself from Bluetooth stalking and more
• How to track anyone via the Find My network
• How to stop being tracked via Bluetooth beacons like AirTag
• How smartphones build a dossier on you
• Why data brokers build dossiers on you, and how to stop them doing so

Millions of IT systems — some of them industrial and IoT — may start behaving unpredictably on January 19. Potential failures include: glitches in processing card payments; false alarms from security systems; incorrect operation of medical equipment; failures in automated lighting, heating, and water supply systems; and many more or less serious types of errors. The catch is — it will happen on January 19, 2038. Not that that’s a reason to relax — the time left to prepare may already be insufficient. The cause of this mass of problems will be an overflow in the integers storing date and time. While the root cause of the error is simple and clear, fixing it will require extensive and systematic efforts on every level — from governments and international bodies and down to organizations and private individuals.

The unwritten standard of the Unix epoch

The Unix epoch is the timekeeping system adopted by Unix operating systems, which became popular across the entire IT industry. It counts the seconds from 00:00:00 UTC on January 1, 1970, which is considered the zero point. Any given moment in time is represented as the number of seconds that have passed since that date. For dates before 1970, negative values are used. This approach was chosen by Unix developers for its simplicity — instead of storing the year, month, day, and time separately, only a single number is needed. This facilitates operations like sorting or calculating the interval between dates. Today, the Unix epoch is used far beyond Unix systems: in databases, programming languages, network protocols, and in smartphones running iOS and Android.

The Y2K38 time bomb

Initially, when Unix was developed, a decision was made to store time as a 32-bit signed integer. This allowed for representing a date range from roughly 1901 to 2038. The problem is that on January 19, 2038, at 03:14:07 UTC, this number will reach its maximum value (2,147,483,647 seconds) and overflow, becoming negative, and causing computers to “teleport” from January 2038 back to December 13, 1901. In some cases, however, shorter “time travel” might happen — to point zero, which is the year 1970.

This event, known as the “year 2038 problem”, “Epochalypse”, or “Y2K38”, could lead to failures in systems that still use 32-bit time representation — from POS terminals, embedded systems, and routers, to automobiles and industrial equipment. Modern systems solve this problem by using 64 bits to store time. This extends the date range to hundreds of billions of years into the future. However, millions of devices with 32-bit dates are still in operation, and will require updating or replacement before “day Y” arrives.

In this context, 32 and 64 bits refer specifically to the date storage format. Just because an operating system or processor is 32-bit or 64-bit, it doesn’t automatically mean it stores the date in its “native” bit format. Furthermore, many applications store dates in completely different ways, and might be immune to the Y2K38 problem, regardless of their bitness.

In cases where there’s no need to handle dates before 1970, the date is stored as an unsigned 32-bit integer. This type of number can represent dates from 1970 to 2106, so the problem will arrive in the more distant future.

Differences from the year 2000 problem

The infamous year 2000 problem (Y2K) from the late 20^th century was similar in that systems storing the year as two digits could mistake the new date for the year 1900. Both experts and the media feared a digital apocalypse, but in the end there were just numerous isolated manifestations that didn’t lead to global catastrophic failures.

The key difference between Y2K38 and Y2K is the scale of digitization in our lives. The number of systems that will need updating is way higher than the number of computers in the 20^th century, and the count of daily tasks and processes managed by computers is beyond calculation. Meanwhile, the Y2K38 problem has already been, or will soon be, fixed in regular computers and operating systems with simple software updates. However, the microcomputers that manage air conditioners, elevators, pumps, door locks, and factory assembly lines could very well chug along for the next decade with outdated, Y2K38-vulnerable software versions.

Potential problems of the Epochalypse

The date’s rolling over to 1901 or 1970 will impact different systems in different ways. In some cases, like a lighting system programmed to turn on every day at 7pm, it might go completely unnoticed. In other systems that rely on complete and accurate timestamps, a full failure could occur — for example, in the year 2000, payment terminals and public transport turnstiles stopped working. Comical cases are also possible, like issuing a birth certificate with a date in 1901. Far worse would be the failure of critical systems, such as a complete shutdown of a heating system, or the failure of a bone marrow analysis system in a hospital.

Cryptography holds a special place in the Epochalypse. Another crucial difference between 2038 and 2000 is the ubiquitous use of encryption and digital signatures to protect all communications. Security certificates generally fail verification if the device’s date is incorrect. This means a vulnerable device would be cut off from most communications — even if its core business applications don’t have any code that incorrectly handles the date.

Unfortunately, the full spectrum of consequences can only be determined through controlled testing of all systems, with separate analysis of a potential cascade of failures.

The malicious exploitation of Y2K38

IT and InfoSec teams should treat Y2K38 not as a simple software bug, but as a vulnerability that can lead to various failures, including denial of service. In some cases, it can even be exploited by malicious actors. To do this, they need the ability to manipulate the time on the targeted system. This is possible in at least two scenarios:

• Interfering with NTP protocol data by feeding the attacked system a fake time server
• Spoofing the GPS signal — if the system relies on satellite time

Exploitation of this error is most likely in OT and IoT systems, where vulnerabilities are traditionally slow to be patched, and the consequences of a failure can be far more substantial.

An example of an easily exploitable vulnerability related to time counting is CVE-2025-55068 (CVSSv3 8.2, CVSSv4 base 8.8) in Dover ProGauge MagLink LX4 automatic fuel-tank gauge consoles. Time manipulation can cause a denial of service at the gas station, and block access to the device’s web management panel. This defect earned its own CISA advisory.

The current status of Y2K38 mitigation

The foundation for solving the Y2K38 problem has been successfully laid in major operating systems. The Linux kernel added support for 64-bit time even on 32-bit architectures starting with version 5.6 in 2020, and 64-bit Linux was always protected from this issue. The BSD family, macOS, and iOS use 64-bit time on all modern devices. All versions of Windows released in the 21^st century aren’t susceptible to Y2K38.

The situation at the data storage and application level is far more complex. Modern file systems like ZFS, F2FS, NTFS, and ReFS were designed with 64-bit timestamps, while older systems like ext2 and ext3 remain vulnerable. Ext4 and XFS require specific flags to be enabled (extended inode for ext4, and bigtime for XFS), and might need offline conversion of existing filesystems. In the NFSv2 and NFSv3 protocols, the outdated time storage format persists. It’s a similar patchwork landscape in databases: the TIMESTAMP type in MySQL is fundamentally limited to the year 2038, and requires migration to DATETIME, while the standard timestamp types in PostgreSQL are safe. For applications written in C, pathways have been created to use 64-bit time on 32-bit architectures, but all projects require recompilation. Languages like Java, Python, and Go typically use types that avoid the overflow, but the safety of compiled projects depends on whether they interact with vulnerable libraries written in C.

A massive number of 32-bit systems, embedded devices, and applications remain vulnerable until they’re rebuilt and tested, and then have updates installed by all their users.

Various organizations and enthusiasts are trying to systematize information on this, but their efforts are fragmented. Consequently, there’s no “common Y2K38 vulnerability database” out there (1, 2, 3, 4, 5).

Approaches to fixing Y2K38

The methodologies created for prioritizing and fixing vulnerabilities are directly applicable to the year 2038 problem. The key challenge will be that no tool today can create an exhaustive list of vulnerable software and hardware. Therefore, it’s essential to update inventory of corporate IT assets, ensure that inventory is enriched with detailed information on firmware and installed software, and then systematically investigate the vulnerability question.

The list can be prioritized based on the criticality of business systems and the data on the technology stack each system is built on. The next steps are: studying the vendor’s support portal, making direct inquiries to hardware and software manufacturers about their Y2K38 status, and, as a last resort, verification through testing.

When testing corporate systems, it’s critical to take special precautions:

• Never test production systems.
• Create a data backup immediately before the test.
• Isolate the system being tested from communications so it can’t confuse other systems in the organization.
• If changing the date uses NTP or GPS, ensure the 2038 test signals cannot reach other systems.
• After testing, set the systems back to the correct time, and thoroughly document all observed system behaviors.

If a system is found to be vulnerable to Y2K38, a fixing timeline should be requested from the vendor. If a fix is impossible, plan a migration; fortunately, the time we have left still allows for updating even fairly complex and expensive systems.

The most important thing in tackling Y2K38 is not to think of it as a distant future problem whose solution can easily wait another five to eight years. It’s highly likely that we already have insufficient time to completely eradicate the defect. However, within an organization and its technology fleet, careful planning and a systematic approach to solving the problem will allow to actually make it in time.

Brand, website, and corporate mailout impersonation is becoming an increasingly common technique used by cybercriminals. The World Intellectual Property Organization (WIPO) reported a spike in such incidents in 2025. While tech companies and consumer brands are the most frequent targets, every industry in every country is generally at risk. The only thing that changes is how the imposters exploit the fakes In practice, we typically see the following attack scenarios:

• Luring clients and customers to a fake website to harvest login credentials for the real online store, or to steal payment details for direct theft.
• Luring employees and business partners to a fake corporate login portal to acquire legitimate credentials for infiltrating the corporate network.
• Prompting clients and customers to contact the scammers under various pretexts: getting tech support, processing a refund, entering a prize giveaway, or claiming compensation for public events involving the brand. The goal is to then swindle the victims out of as much money as possible.
• Luring business partners and employees to specially crafted pages that mimic internal company systems, to get them to approve a payment or redirect a legitimate payment to the scammers.
• Prompting clients, business partners, and employees to download malware — most often an infostealer — disguised as corporate software from a fake company website.

The words “luring” and “prompting” here imply a whole toolbox of tactics: email, messages in chat apps, social media posts that look like official ads, lookalike websites promoted through SEO tools, and even paid ads.

These schemes all share two common features. First, the attackers exploit the organization’s brand, and strive to mimic its official website, domain name, and corporate style of emails, ads, and social media posts. And the forgery doesn’t have to be flawless — just convincing enough for at least some of business partners and customers. Second, while the organization and its online resources aren’t targeted directly, the impact on them is still significant.

Business damage from brand impersonation

When fakes are crafted to target employees, an attack can lead to direct financial loss. An employee might be persuaded to transfer company funds, or their credentials could be used to steal confidential information or launch a ransomware attack.

Attacks on customers don’t typically imply direct damage to the company’s coffers, but they cause substantial indirect harm in the following areas:

• Strain on customer support. Customers who “bought” a product on a fake site will likely bring their issues to the real customer support team. Convincing them that they never actually placed an order is tough, making each case a major time waster for multiple support agents.
• Reputational damage. Defrauded customers often blame the brand for failing to protect them from the scam, and also expect compensation. According to a European survey, around half of affected buyers expect payouts and may stop using the company’s services — often sharing their negative experience on social media. This is especially damaging if the victims include public figures or anyone with a large following.
• Unplanned response costs. Depending on the specifics and scale of an attack, an affected company might need digital forensics and incident response (DFIR) services, as well as consultants specializing in consumer law, intellectual property, cybersecurity, and crisis PR.
• Increased insurance premiums. Companies that insure businesses against cyber-incidents factor in fallout from brand impersonation. An increased risk profile may be reflected in a higher premium for a business.
• Degraded website performance and rising ad costs. If criminals run paid ads using a brand’s name, they siphon traffic away from its official site. Furthermore, if a company pays to advertise its site, the cost per click rises due to the increased competition. This is a particularly acute problem for IT companies selling online services, but it’s also relevant for retail brands.
• Long-term metric decline. This includes drops in sales volume, market share, and market capitalization. These are all consequences of lost trust from customers and business partners following major incidents.

Does insurance cover the damage?

Popular cyber-risk insurance policies typically only cover costs directly tied to incidents explicitly defined in the policy — think data loss, business interruption, IT system compromise, and the like. Fake domains and web pages don’t directly damage a company’s IT systems, so they’re usually not covered by standard insurance. Reputational losses and the act of impersonation itself are separate insurance risks, requiring expanded coverage for this scenario specifically.

Of the indirect losses we’ve listed above, standard insurance might cover DFIR expenses and, in some cases, extra customer support costs (if the situation is recognized as an insured event). Voluntary customer reimbursements, lost sales, and reputational damage are almost certainly not covered.

What to do if your company is attacked by clones

If you find out someone is using your brand’s name for fraud, it makes sense to do the following:

• Send clear, straightforward notifications to your customers explaining what happened, what measures are being taken, and how to verify the authenticity of official websites, emails, and other communications.
• Create a simple “trust center” page listing your official domains, social media accounts, app store links, and support contacts. Make it easy to find and keep it updated.
• Monitor new registrations of social media pages and domain names that contain your brand names to spot the clones before an attack kicks off.
• Follow a takedown procedure. This involves gathering evidence, filing complaints with domain registrars, hosting providers, and social media administrators, then tracking the status until the fakes are fully removed. For a complete and accurate record of violations, preserve URLs, screenshots, metadata, and the date and time of discovery. Ideally, also examine the source code of fake pages, as it might contain clues pointing to other components of the criminal operation.
• Add a simple customer reporting form for suspicious sites or messages to your official website and/or branded app. This helps you learn about problems early.
• Coordinate activities between your legal, cybersecurity, and marketing teams. This ensures a consistent, unified, and effective response.

How to defend against brand impersonation attacks

While the open nature of the internet and the specifics of these attacks make preventing them outright impossible, a business can stay on top of new fakes and have the tools ready to fight back.

• Continuously monitor for suspicious public activity using specialized monitoring services. The most obvious indicator is the registration of domains similar to your brand name, but there are others — like someone buying databases related to your organization on the dark web. Comprehensive monitoring of all platforms is best outsourced to a specialized service provider, such as Kaspersky Digital Footprint Intelligence (DFI).
• The quickest and simplest way to take down a fake website or social media profile is to file a trademark infringement complaint. Make sure your portfolio of registered trademarks is robust enough to file complaints under UDRP procedures before you need it.
• When you discover fakes, deploy UDRP procedures promptly to have the fake domains transferred or removed. For social media, follow the platform’s specific infringement procedure — easily found by searching for “[social media name] trademark infringement” (for example, “LinkedIn trademark infringement”). Transferring the domain to the legitimate owner is preferred over deletion, as it prevents scammers from simply re-registering it. Many continuous monitoring services, such as Kaspersky Digital Footprint Intelligence, also offer a rapid takedown service, filing complaints on the protected brand’s behalf.
• Act quickly to block fake domains on your corporate systems. This won’t protect partners or customers, but it’ll throw a wrench into attacks targeting your own employees.
• Consider proactively registering your company’s website name and common variations (for example, with and without hyphens) in all major top-level domains, such as .com, and local extensions. This helps protect partners and customers from common typos and simple copycat sites.

Thanks to the convenience of NFC and smartphone payments, many people no longer carry wallets or remember their bank card PINs. All their cards reside in a payment app, and using that is quicker than fumbling for a physical card. Mobile payments are also secure — the technology was developed relatively recently and includes numerous anti-fraud protections. Still, criminals have invented several ways to abuse NFC and steal your money. Fortunately, protecting your funds is straightforward: just know about these tricks and avoid risky NFC usage scenarios.

What are NFC relay and NFCGate?

NFC relay is a technique where data wirelessly transmitted between a source (like a bank card) and a receiver (like a payment terminal) is intercepted by one intermediate device, and relayed in real time to another. Imagine you have two smartphones connected via the internet, each with a relay app installed. If you tap a physical bank card against the first smartphone and hold the second smartphone near a terminal or ATM, the relay app on the first smartphone will read the card’s signal using NFC, and relay it in real time to the second smartphone, which will then transmit this signal to the terminal. From the terminal’s perspective, it all looks like a real card is tapped on it — even though the card itself might physically be in another city or country.

This technology wasn’t originally created for crime. The NFCGate app appeared in 2015 as a research tool after it was developed by students at the Technical University of Darmstadt in Germany. It was intended for analyzing and debugging NFC traffic, as well as for education purposes and experiments with contactless technology. NFCGate was distributed as an open-source solution and used in academic and enthusiast circles.

Five years later, cybercriminals caught on to the potential of NFC relay and began modifying NFCGate by adding mods that allowed it to run through a malicious server, disguise itself as legitimate software, and perform social engineering scenarios.

What began as a research project morphed into the foundation for an entire class of attacks aimed at draining bank accounts without physical access to bank cards.

A history of misuse

The first documented attacks using a modified NFCGate occurred in late 2023 in the Czech Republic. By early 2025, the problem had become large scale  and noticeable: cybersecurity analysts uncovered more than 80 unique malware samples built on the NFCGate framework. The attacks evolved rapidly, with NFC relay capabilities being integrated into other malware components.

By February 2025, malware bundles combining CraxsRAT and NFCGate emerged, allowing attackers to install and configure the relay with minimal victim interaction. A new scheme, a so-called “reverse” version of NFCGate, appeared in spring 2025, fundamentally changing the attack’s execution.

Particularly noteworthy is the RatOn Trojan, first detected in the Czech Republic. It combines remote smartphone control with NFC relay capabilities, letting attackers target victims’ banking apps and cards through various technique combinations. Features like screen capture, clipboard data manipulation, SMS sending, and stealing info from crypto wallets and banking apps give criminals an extensive arsenal.

Cybercriminals have also packaged NFC relay technology into malware-as-a-service (MaaS) offerings, and reselling them to other threat actors through subscription. In early 2025, analysts uncovered a new and sophisticated Android malware campaign in Italy, dubbed SuperCard X. Attempts to deploy SuperCard X were recorded in Russia in May 2025, and in Brazil in August of the same year.

The direct NFCGate attack

The direct attack is the original criminal scheme exploiting NFCGate. In this scenario, the victim’s smartphone plays the role of the reader, while the attacker’s phone acts as the card emulator.

First, the fraudsters trick the user into installing a malicious app disguised as a banking service, a system update, an “account security” app, or even a popular app like TikTok. Once installed, the app gains access to both NFC and the internet — often without requesting dangerous permissions or root access. Some versions also ask for access to Android accessibility features.

Then, under the guise of identity verification, the victim is prompted to tap their bank card to their phone. When they do, the malware reads the card data via NFC and immediately sends it to the criminals’ server. From there, the information is relayed to a second smartphone held by a money mule, who helps extract the money. This phone then emulates the victim’s card to make payments at a terminal or withdraw cash from an ATM.

The fake app on the victim’s smartphone also asks for the card PIN — just like at a payment terminal or ATM — and sends it to the attackers.

In early versions of the attack, criminals would simply stand ready at an ATM with a phone to use the duped user’s card in real time. Later, the malware was refined so the stolen data could be used for in-store purchases in a delayed, offline mode, rather than in a live relay.

For the victim, the theft is hard to notice: the card never left their possession, they didn’t have to manually enter or recite its details, and the bank alerts about the withdrawals can be delayed or even intercepted by the malicious app itself.

Among the red flags that should make you suspect a direct NFC attack are:

• prompts to install apps not from official stores;
• requests to tap your bank card on your phone.

The reverse NFCGate attack

The reverse attack is a newer, more sophisticated scheme. The victim’s smartphone no longer reads their card — it emulates the attacker’s card. To the victim, everything appears completely safe: there’s no need to recite card details, share codes, or tap a card to the phone.

Just like with the direct scheme, it all starts with social engineering. The user gets a call or message convincing them to install an app for “contactless payments”, “card security”, or even “using central bank digital currency”. Once installed, the new app asks to be set as the default contactless payment method — and this step is critically important. Thanks to this, the malware requires no root access — just user consent.

The malicious app then silently connects to the attackers’ server in the background, and the NFC data from a card belonging to one of the criminals is transmitted to the victim’s device. This step is completely invisible to the victim.

Next, the victim is directed to an ATM. Under the pretext of “transferring money to a secure account” or “sending money to themselves”, they are instructed to tap their phone on the ATM’s NFC reader. At this moment, the ATM is actually interacting with the attacker’s card. The PIN is dictated to the victim beforehand — presented as “new” or “temporary”.

The result is that all the money deposited or transferred by the victim ends up in the criminals’ account.

The hallmarks of this attack are:

• requests to change your default NFC payment method;
• a “new” PIN;
• any scenario where you’re told to go to an ATM and perform actions there under someone else’s instructions.

How to protect yourself from NFC relay attacks

NFC relay attacks rely not so much on technical vulnerabilities as on user trust. Defending against them comes down to some simple precautions.

• Make sure you keep your trusted contactless payment method (like Google Pay or Samsung Pay) as the default.
• Never tap your bank card on your phone at someone else’s request, or because an app tells you to. Legitimate apps might use your camera to scan a card number, but they’ll never ask you to use the NFC reader for your own card.
• Never follow instructions from strangers at an ATM — no matter who they claim to be.
• Avoid installing apps from unofficial sources. This includes links sent via messaging apps, social media, SMS, or recommended during a phone call — even if they come from someone claiming to be customer support or the police.
• Use comprehensive security on your Android smartphones to block scam calls, prevent visits to phishing sites, and stop malware installation.
• Stick to official app stores only. When downloading from a store, check the app’s reviews, number of downloads, publication date, and rating.
• When using an ATM, rely on your physical card instead of your smartphone for the transaction.
• Make it a habit to regularly check the “Payment default” setting in your phone’s NFC menu. If you see any suspicious apps listed, remove them immediately and run a full security scan on your device.
• Review the list of apps with accessibility permissions — this is a feature commonly abused by malware. Either revoke these permissions for any suspicious apps, or uninstall the apps completely.
• Save the official customer service numbers for your banks in your phone’s contacts. At the slightest hint of foul play, call your bank’s hotline directly without delay.
• If you suspect your card details may have been compromised, block the card immediately.