Key Takeaways:

• Traditional vulnerability management tools can no longer keep up with the speed of modern exploitation—threat context is now mandatory.
• Threat and Vulnerability Management (TVM) systems unify asset discovery, vulnerability data, and real-time external threat intelligence to prioritize real risk.
• Static CVSS scores fail to reflect exploitation likelihood; intelligence-driven, dynamic risk scoring is essential in 2026.
• Organizations that integrate vulnerability intelligence and attack surface intelligence reduce remediation time and security waste, enhancing detection and remediation while reducing alert fatigue.

Why Threat and Vulnerability Management Must Evolve in 2026

Security teams currently find themselves at a crossroads. Year over year, CVE volumes continue to surge higher and higher. Exploitation is faster, more automated, and more targeted, meaning attacks are growing in volume, velocity, and sophistication alike. As a result, security teams are expected to “patch faster” with fewer resources and can no longer realistically keep up with this ever-rising tide of threats.

Thanks to these forces, security teams have found themselves in a state of affairs in which vulnerability management has become an exercise in sheer volume, not risk. Day in and day out, teams are overwhelmed by alerts that lack real-world context, making it all but impossible to assess the actual degree of risk.

Thankfully, there is a solution. Threat-informed vulnerability management (TVM) has emerged to counteract this trend, enabling security teams to intelligently address weaponized vulnerabilities, zero-day exploits, and supply chain and cloud-native risk. All this comes along with much-needed relief from creeping alert-fatigue.

In 2026, effective cybersecurity programs will be defined not by how many vulnerabilities they detect but by how precisely they understand, prioritize, and neutralize real threats using intelligence-driven TVM systems.

The Core Problem: Alert Fatigue and Prioritization Failure

As it stands today, the explosion in disclosed vulnerabilities (CVEs) has outpaced humans’ abilities to triage and manage patching effectively. Today, the vast majority of organizations are incapable of remediating more than a fraction of the total identified issues affecting the ecosystem.

Traditionally, using a standard CVSS (Common Vulnerability Scoring System) was enough to overcome these challenges of prioritization. CVSS is an open, standardized framework used to assess the severity of security vulnerabilities by assigning a numerical score based on factors like exploitability, impact, and scope. Organizations use CVSS scores to prioritize remediation and compare vulnerabilities consistently across systems and vendors.

However, CVSS only measures theoretical severity, not exploitation likelihood. It misses critical pieces of context for prioritization decisions such as:

• Is exploit code available?
• Is the vulnerability actively exploited?
• Are threat actors discussing or operationalizing it?

As a result, high-severity CVEs that pose little real-world risk continue to consume time and resources, leading us back once again to the issue of alert fatigue and the inability to effectively triage and patch the most pressing vulnerabilities.

At the same time, we are seeing modern organizations struggle with a “silo problem,” in which security, IT, and CTI (cyber threat intelligence) teams operate independently and with limited visibility and collaboration between one another. In many organizations, each of these teams ends up using different tools, establishing different priorities, sharing findings infrequently if at all, and adopting entirely different “risk languages” through which they understand, prioritize, and address threats.

Taken broadly, this leaves organizations woefully lacking a unified, intelligence-driven view of risk. Without this, many adopt a de facto policy of “patch everything”. And it comes with significant costs, including:

• Operational drag and burnout
• Delayed remediation of truly dangerous vulnerabilities
• Increased business risk despite increased effort
• Fractured security operations

Both individually, and in the aggregate, these side-effects come at a significant detriment to organizational security. And as the number and diversity of CVEs continues to expand, the greater that cost becomes. Moving forward, organizations must find a better way.

The Evolving Threat Landscape Demands a New Approach

Today’s ever-changing landscape means that organizations must evolve along with it or risk falling dangerously behind. The rise of rapidly weaponized vulnerabilities (i.e., known software weaknesses that have moved beyond disclosure and into active attacker use) reflects a fundamental shift in how quickly and deliberately adversaries turn CVEs into operational threats. Today, the gap between disclosure, proof-of-concept release, and active exploitation has collapsed from months to days (or even hours), driven largely by exploit marketplaces, automated scanning, and widely shared tooling.

Attackers increasingly prioritize vulnerabilities that are easy to exploit, broadly applicable across cloud services, edge devices, and common dependencies, and capable of delivering fast returns. Once weaponized, these vulnerabilities manifest not as theoretical risk but as active intrusion campaigns, ransomware operations, and opportunistic internet-wide exploitation, making threat context essential for distinguishing true danger from background noise.

At the same time that weaponization is accelerating, attack surfaces are expanding. The average attack surface today is expanding and fragmenting across hybrid and multi-cloud environments, all of which is worsened by SaaS sprawl, shadow IT, and third-party and supply chain exposure. In this environment, it is absolutely critical that security teams have a clear understanding of vulnerabilities vs. threats, and work to establish an integrated approach between the two.

In short, a vulnerability is a technical weakness, while a threat is an actor, campaign or event at work exploiting that weakness. In order to be truly effective, modern threat vulnerability management (TVM) systems must merge both concepts to reflect real risk and separate signal from noise.

What Is Threat and Vulnerability Management (TVM)?

Threat and Vulnerability Management (TVM) — also called Threat-Informed Vulnerability Management — is a continuous, intelligence-driven process that prioritizes remediation based on three core variables:

• Active exploitation
• Threat actor behavior
• Asset criticality

TVM differs from traditional vulnerability management (VM) in a number of critical ways. Traditional VM relies on periodic scans, static severity scoring, and a largely reactive patching process. TVM, on the other hand, employs continuous monitoring, external threat intelligence enrichment, and close-loop remediation and validation.

This continuous, context-rich approach is foundational for modern security programs. Rather than inundating security teams with decontextualized CVEs and indiscriminate patching, modern TVM systems align security efforts with attacker reality. Reactive patching is replaced with proactive, risk-based decision-making, and as a result, organizations are able to reduce noise while simultaneously increasing the impact of their security operations.

The Five Core Pillars of Modern TVM Systems

As the speed and breadth of today’s threats continue to grow, traditional VM, being fundamentally reactive in nature, is no longer enough to keep up. In a world where vulnerabilities are exposed by the day, TVM offers much-needed efficiency, intelligence, and proactiveness. However, not all TVM systems are created equally. Here are five core pillars of effective modern TVM systems to help you evaluate and assess solutions on the market.

1. Continuous Asset Discovery & Inventory

Modern TVM systems are invaluable in that they provide full visibility across the entirety of an organization’s growing and fragmented attack surface. This includes external-facing assets, shadow IT, and cloud and SaaS environments alike. By providing continuous asset discovery and a timely, up-to-date inventory of one’s assets, TVM systems allow for real-time, comprehensive, attack-surface management.

Remember, you can’t defend what you can’t see. That’s why attack surface management (ASM) is a prerequisite for effective TVM. Without accurate, up-to-date asset inventories, vulnerability data is incomplete and misleading. Continuous discovery ensures defenders see their environment the way attackers do.

2. Vulnerability Assessment & Scoring

TVM goes beyond internal scanning tools to identify vulnerabilities exposed to the internet and reassess them continuously as environments change. This includes tracking misconfigurations, outdated services, and newly introduced exposure, not just known CVEs.

3. External Threat Context Enrichment

This is where TVM fundamentally diverges from legacy approaches. External threat intelligence enriches vulnerability data with insight from dark web and criminal forums, exploit marketplaces, malware telemetry, and active attack campaigns.

Vulnerabilities are mapped to known threat actors, active exploitation, and MITRE ATT&CK® techniques, ultimately transforming raw findings into actionable intelligence.

4. Risk-Based Prioritization (RBVM)

Risk-based vulnerability management prioritizes issues based on the probability of exploitation, asset importance, and threat actor interest. This shifts the focus from “most severe” to “most dangerous,” enabling teams to address the vulnerabilities that pose the greatest immediate risk to their organizations.

5. Automated Remediation & Verification

Modern TVM integrates directly with IT and SecOps workflows, pushing prioritized findings into ticketing and automation platforms. Just as importantly, it verifies remediation to confirm that patches were applied and exposure was actually reduced, creating a continuous feedback loop.

These five pillars of effective TVM systems come together to create a whole that is greater than the sum of its parts. These systems, unlike their predecessors, are designed to continuously monitor and triage real threats and vulnerabilities in context and ensure awareness and proactive mitigation without the risk of burn-out and alert fatigue.

Stop Patching Everything — Use Intelligence to Prioritize Real Risk

The scale of the CVE problem is overwhelming. Tens of thousands of vulnerabilities are disclosed each year, yet only a small fraction are ever exploited in the wild. Treating them all as equally urgent is not just inefficient — it’s dangerous.

Vulnerability intelligence changes the equation by tracking a CVE across its full lifecycle, from initial disclosure to weaponization, exploitation, and criminal adoption. This enables dynamic risk scoring that reflects real-world conditions rather than static assumptions.

Dynamic risk scoring incorporates evidence of active exploitation, availability of exploit code, dark web chatter, and threat actor interest. As conditions change, so does the risk score, ensuring prioritization remains aligned with attacker behavior.

The operational impact is significant. Security teams can focus remediation on the top 1% of vulnerabilities that pose immediate risk, respond faster, reduce operational cost, and strengthen overall security posture.

See Your Risk Like an Attacker: The Full Attack Surface View

In today’s threat landscape, security teams must recast the way they envision their roles. Rather than operating in a reactive, defensive manner at all times, security teams should think more like their adversaries, taking a complete view of their attack surface and leveraging modern tools and technologies to ensure intelligent, prioritized defenses. The following three key concepts will help you take on that mentality.

1. The Visibility Gap: Unknown assets create unknown risk. Traditional scanners often miss orphaned domains, misconfigured cloud services, and forgotten infrastructure — precisely the assets attackers look for first.
2. Attack Surface Intelligence Explained: Attack surface intelligence provides continuous mapping of domains, IPs, cloud assets, and external services. It identifies exposures attackers see before defenders do, enabling proactive remediation rather than reactive cleanup.
3. Connecting the Dots with Vulnerability Tools: When integrated with vulnerability scanners like Qualys and Tenable, attack surface intelligence provides a unified, prioritized view of exposure. Intelligence-driven platforms serve as a single source of truth for risk decisions, enabling teams to connect vulnerabilities to real-world exposure and threat activity.

Three Strategic Recommendations for Security Leaders

Most organizations remain behind the curve in threat and vulnerability management. Knowing what we know now, there are three strategic steps security leaders can take to reclaim control.

1. Bridge the Gap Between Security and IT

Establish a shared, intelligence-driven risk language. Align SLAs with real-world risk rather than raw severity scores, ensuring remediation efforts focus on what matters most.

2. Embrace Automation and Workflow Integration

Push prioritized findings directly into platforms like ServiceNow and SOAR tools. Reducing manual handoffs accelerates remediation and minimizes delays.

3. Measure What Matters — Time-to-Remediate (TTR)

Shift KPIs toward time-to-remediate actively exploited vulnerabilities and reduction in exposure windows. These metrics demonstrate real ROI and security impact.

The Path Forward Is Threat-Informed: Strengthen Your Threat and Vulnerability Strategy

Volume-based vulnerability management is no longer viable. As we progress through 2026, threat context is not optional. It is foundational.

Future-ready security programs are intelligence-led, automation-enabled, and attacker-aware. Recorded Future sits at the center of this shift, providing the intelligence backbone required to move from reactive patching to proactive risk reduction.

Explore how Recorded Future Vulnerability Intelligence and Attack Surface Intelligence can help your organization transition from alert-driven vulnerability management to intelligence-driven risk reduction.

By unifying threat intelligence, vulnerability data, and attack surface visibility, organizations can reduce alert fatigue, prioritize what truly matters, and proactively harden defenses against real-world threats before attackers exploit them.

Key Takeaways

• Effective ransomware detection requires three complementary layers: endpoint and extended detection and response (EDR/XDR) to monitor device-level activity, network detection and response (NDR) to catch lateral movement, and threat intelligence tools to provide context that enables efficient prioritization.
• The most valuable detection happens before ransomware encryption begins. Tools must identify precursor behaviors like reconnaissance, credential theft, and data staging rather than waiting for known indicators of compromise.
• Intelligence quality determines detection quality: even sophisticated security tools require real-time threat data about active ransomware campaigns, attacker infrastructure, and current tactics, techniques, and procedures (TTPs) to distinguish genuine threats from noise.
• Recorded Future strengthens the entire detection stack by providing organization-specific threat intelligence, early detection capabilities (in some cases, identifying victims up to 30 days before public extortion), and vulnerability intelligence focused on what ransomware groups are actively exploiting.

Introduction

The ransomware playbook has fundamentally changed. Instead of casting wide nets with opportunistic phishing campaigns, attackers now focus on big-game hunting: targeting high-value enterprises with data theft and double or triple extortion tactics. Threat actors purchase pre-compromised access from brokers, exploit newly disclosed vulnerabilities within hours, and use automation to compress weeks-long campaigns into days.

The results are stark. Ransomware now appears in 44% of breaches, up from 32% the prior year, according to the 2025 Verizon Data Breach Investigations Report. Traditional signature-based detection tools often can't keep pace because ransomware groups continuously rotate their infrastructure, modify malware variants, and adopt new tactics faster than defenses can update. By the time a signature is written, the threat has already evolved.

This gap has created demand for a different approach: intelligence-driven ransomware detection. Rather than waiting for known indicators of compromise, these tools identify the precursor behaviors that happen before encryption (e.g. reconnaissance, credential theft, lateral movement, privilege escalation, and data staging).

The key is continuous external intelligence that maps what's happening in your environment to active campaigns and specific ransomware families operating in the wild.

The most effective defense combines three layers: endpoint and extended detection and response (EDR/XDR) to catch suspicious behaviors on devices, network detection and response (NDR) with deception technology to spot lateral movement, and threat intelligence tools that provide the real-time context tying it all together. When these tools share a common intelligence foundation, they can reveal malicious intent well before encryption begins.

The Ransomware Detection Tool Landscape: Three Pillars of Defense

Effective ransomware detection generally requires three complementary tool categories, each targeting different stages of an attack.

1. Endpoint and Extended Detection and Response (EDR/XDR) Tools

EDR and XDR platforms form the first line of defense, monitoring individual devices and user activity for signs of compromise.

Core Functionality

EDR and XDR solutions monitor endpoints for suspicious behaviors like privilege escalation, credential dumping, unusual process creation, and bulk file modifications. When they detect threats, these tools automatically isolate devices, roll back changes, and contain threats, cutting response time from hours to seconds.

How Threat Intelligence Enhances EDR/XDR

Threat intelligence connects endpoint activity to active campaigns in the wild. When an EDR tool flags suspicious activity, intelligence context reveals whether it matches known campaigns from groups like LockBit, ALPHV/BlackCat, or BlackBasta. This can dramatically reduce false positives by distinguishing unusual-but-legitimate administrative work from activity aligned with active ransomware operations.

Example Tools

• CrowdStrike Falcon delivers strong behavioral detection capabilities tied to comprehensive actor profiling. The platform's threat graph continuously correlates endpoint telemetry with global threat intelligence, enabling rapid identification of ransomware precursors.
• Microsoft Defender XDR integrates telemetry across identity systems, endpoints, email, and cloud applications. This unified visibility helps security teams identify cross-domain attack patterns that indicate ransomware preparation, such as credential theft followed by lateral movement.
• SentinelOne employs behavioral AI to detect malicious activity and offers automated rollback features that can reverse ransomware encryption and file modifications, effectively restoring systems to their pre-attack state.

2. Network Detection and Response (NDR) Tools

While EDR focuses on individual endpoints, NDR tools monitor the network layer to catch attackers as they move between systems.

Core Functionality

NDR platforms watch internal network traffic to catch attackers moving laterally, scanning for targets, or accessing resources they shouldn't. The more advanced versions include deception technology like honeypots, fake credentials, and decoy systems that look like attractive targets. When attackers interact with these decoys during reconnaissance, security teams get early warnings before any real damage occurs.

How Threat Intelligence Improves NDR and Deception

Threat intelligence helps organizations customize deception environments based on active ransomware groups in their industry. When NDR tools spot anomalies such as unusual file sharing, unexpected queries, or abnormal transfers, intelligence matches these to current attack techniques, distinguishing administrative work from reconnaissance patterns before data staging begins.

Example Tools

• Vectra AI specializes in detecting lateral movement and privilege misuse by correlating network behaviors with active attacker tradecraft. The platform's AI-driven detection identifies subtle deviations from normal network patterns that indicate ransomware reconnaissance.
• ExtraHop Reveal(x) provides real-time network visibility that identifies reconnaissance activity and command-and-control (C2) communications. The platform's deep packet inspection capabilities reveal malicious traffic even when encrypted or obfuscated.
• Illusive (now part of Zscaler) deploys deception technology specifically tuned to adversary behaviors. The platform's decoys and fake credentials create a minefield for attackers, triggering high-confidence alerts when threat actors interact with deception assets.

3. Threat Intelligence Tools

The third pillar provides the context that makes endpoint and network detection tools more accurate and actionable.

Core Functionality

Threat intelligence tools pull together global threat data from sources like dark web forums, malware repositories, scanning activity, and criminal infrastructure. They enrich alerts from your other security tools with context about who's behind an attack, which campaign it's part of, and what techniques the attackers are likely to use next.

How Threat Intelligence Strengthens Ransomware Detection

These tools deliver several critical capabilities that transform how security teams identify and respond to ransomware threats:

• Threat Mapping: Identifies whether your organization matches the targeting profile of active ransomware groups based on your industry, size, region, and technology stack. Specific operators are mapped using their TTPs to determine the intent and opportunity of carrying out a successful attack against your business.
• Infrastructure Tracking: Monitors ransomware operators' continuous infrastructure shifts in real-time, identifying new C2 servers, drop sites, and payment infrastructure as they emerge.
• Variant Identification: Rapidly analyzes and disseminates indicators when ransomware groups release new malware variants, enabling detection before signature-based systems receive updates.
• Exploitation Intelligence: Identifies specific CVEs and misconfigurations that attackers are actively weaponizing, moving vulnerability management from severity-score-driven to threat-driven prioritization.
• Risk Scoring: Provides real-time scores combining multiple intelligence signals—indicator prevalence, campaign association, TTP alignment—to guide analysts toward genuine threats rather than generic suspicious activity.

Example Tools

• Recorded Future delivers organization-specific threat intelligence powered by The Intelligence Graph and proprietary AI. The platform provides end-to-end visibility into exposures, while research from its Insikt Group enables early detection of ransomware activity, identifying potential victims up to 30 days before public extortion.
• Flashpoint specializes in deep and dark web intelligence, monitoring criminal forums, marketplaces, and chat channels where ransomware operators communicate, recruit, and trade access. This visibility into adversary communities provides early warnings about emerging threats and campaigns.
• Google Threat Intelligence (formerly Mandiant) combines frontline incident response insights with global threat tracking. The platform leverages intelligence from breach investigations to identify ransomware group behaviors and attack patterns as they emerge.

Choosing the Right Ransomware Detection Tools

Security leaders must distinguish between tools that reduce ransomware risk and those that add noise. The most effective tools share several characteristics.

Security leaders should prioritize:

• Pre-encryption visibility: Detect credential misuse, suspicious access, and lateral movement during reconnaissance and preparation phases when interventions are most effective.
• Context-rich alerts: Alerts should include TTPs, infrastructure associations, and known actor activity and explain not just what triggered an alert but why it matters.
• Integration maturity: Smooth data flow into SIEM, SOAR, and existing investigation workflows without creating siloed intelligence or blind spots.
• Operational efficiency: Tools should reduce alert noise, not add to it, decreasing time-to-detection and time-to-response.
• Relevance: Intelligence must map to current campaigns. Generic or stale indicators waste analyst time and create false confidence.
• Scalability: Handle hybrid environments spanning on-premises infrastructure, multiple cloud providers, and remote endpoints without performance degradation.

How Recorded Future Enables Early Ransomware Detection

The quality of threat intelligence directly determines detection effectiveness. Even sophisticated endpoint and network tools require high-fidelity, current threat data to generate value. Security teams have plenty of options for tools; the real challenge is addressing alert fatigue draining analyst time on false positives instead of credible threats.

Recorded Future functions as the continuous intelligence layer strengthening the entire detection stack. Rather than adding another alert-generating tool, it feeds existing security ecosystems with real-time context about ransomware operator behavior.

Real-Time Relevance Through SecOps Intelligence

Every alert that hits your SIEM or endpoint platform gets automatically enriched with real-time risk scores, associated malware and infrastructure, and links to known attacker techniques and campaigns. Security tools can immediately recognize whether an indicator matches an active ransomware operation, cutting triage time from hours to minutes.

Proactive Mitigation Through Vulnerability Intelligence

Recorded Future identifies which vulnerabilities ransomware groups are actually exploiting right now, not just which ones have the highest theoretical severity ratings. This distinction matters because most high-severity vulnerabilities never get exploited in the wild, while some medium-severity vulnerabilities become critical the moment ransomware operators weaponize them.

The platform shows you which vulnerabilities specific ransomware groups are targeting, where exploit code is available, and which vulnerabilities are generating buzz in criminal forums. This lets security teams prioritize patching based on what attackers are actually doing, focusing on the access vectors most likely to result in ransomware incidents.

Victimology and Anticipation

Intelligence about dark web chatter, leak site activity, and victimology patterns reveals which industries, geographies, and technologies are being targeted. When Recorded Future detects increased targeting of specific sectors, SOC analysts can anticipate attack paths, tighten access controls, and implement protective measures before campaigns reach their network.

This closes the gap between reconnaissance and encryption. Most traditional tools don't trigger alerts until ransomware starts encrypting systems, by which point attackers have already stolen data. Intelligence-driven detection can catch the reconnaissance, credential theft, and lateral movement phases that happen first, shifting your response window from reactive damage control to proactive early containment.

Shifting From Reactive Response to Intelligence-Led Prevention

No single tool stops ransomware. The strongest defense is an integrated ecosystem where endpoint detection, network monitoring, and threat analysis platforms work from the same intelligence foundation.

Intelligence elevates these tools from reactive detection to early recognition of adversary behavior during preparation and reconnaissance phases, enabling intervention before ransomware reaches its destructive phase. Organizations that build detection architecture on real-time threat intelligence will adapt as quickly as their adversaries, maintaining effective defenses as the threat landscape evolves.

December 2025 witnessed a dramatic 120% increase in high-impact vulnerabilities, with Recorded Future's Insikt Group® identifying 22 vulnerabilities requiring immediate remediation, up from 10 in November. The month was dominated by widespread exploitation of Meta's React Server Components flaw.

What security teams need to know:

• React2Shell pandemonium: CVE-2025-55182 triggered a global exploitation wave with multiple threat actors deploying diverse malware families
• China-nexus exploitation intensifies: Earth Lamia, Jackpot Panda, and UAT-9686 leveraged critical flaws for espionage operations
• Public exploits proliferate: Eleven of 22 vulnerabilities have proof-of-concept code available, accelerating exploitation timelines
• Legacy vulnerabilities resurface: CISA added 2018-2022 era flaws to its Known Exploited Vulnerabilities (KEV) catalog, highlighting persistent patch gaps

Bottom line: December's surge reflects both new zero-days and renewed interest in legacy vulnerabilities. React2Shell alone demonstrates how quickly modern web frameworks can become global attack vectors.

Quick Reference Table

All 22 vulnerabilities below were actively exploited in December 2025.

Table 1: List of vulnerabilities that were actively exploited in December based on Recorded Future data (Source: Recorded Future)

Key Trends in December 2025

Affected Vendors

• Fortinet continued vulnerability concerns with two critical authentication bypass flaws
• Google faced three vulnerabilities across Android (2) and Chromium (1) platforms
• Microsoft dealt with a Windows kernel use-after-free vulnerability
• Meta experienced the month's most impactful vulnerability with React2Shell
• Additional affected vendors: Array Networks, Gogs, Gladinet, ASUS, Cisco, Apple, SonicWall, WatchGuard, MongoDB, Digiever, Sierra Wireless, OSGeo, RARLAB, D-Link, and OpenPLC

Most Common Weakness Types

• CWE-22 – Path Traversal
• CWE-347 – Improper Verification of Cryptographic Signature
• CWE-416 – Use After Free
• CWE-434 – Unrestricted Upload of File with Dangerous Type
• CWE-787 – Out-of-bounds Write

Threat Actor Activity

React2Shell exploitation dominated December’s CVE activity:

• Threat actors observed to have exploited this vulnerability:
  • China-nexus actors Earth Lamia and Jackpot Panda
  • China-linked clusters UNC6600, UNC6586, UNC6588, UNC6603, and UNC6595
  • North Korea-linked and financially motivated groups
• Observed payloads included EtherRAT, PeerBlight, CowTunnel, ZinFoq, Kaiji variants, Zndoor, RondoDox, MINOCAT, SNOWLIGHT, COMPOOD, HISONIC, ANGRYREBEL.LINUX, and Weaxor ransomware (using a Cobalt Strike stager)
• Infrastructure connections to HiddenOrbit relay infrastructure and GobRAT relay component

Additional activity:

• UAT-9686 exploited Cisco Secure Email Gateway (CVE-2025-20393), deploying AquaShell, AquaPurge, and AquaTunnel
• Unknown actors leveraged Gogs vulnerability (CVE-2025-8110) for Supershell malware deployment

Priority Alert: Active Exploitation

These vulnerabilities demand immediate attention due to confirmed widespread exploitation.

CVE-2025-55182 | Meta React Server Components (React2Shell)

Risk Score: 99 (Very Critical) | CISA KEV: Added December 5, 2025

Why this matters: Unauthenticated RCE affects React and Next.js, among the world's most popular web frameworks. Multiple threat actors are actively exploiting vulnerable instances with diverse malware payloads.

Affected versions:

• React packages: react-server-dom-webpack, react-server-dom-parcel, react-server-dom-turbopack (19.0.0, 19.1.0, 19.1.1, and 19.2.0)
• Next.js: 15.x, 16.x, and Canary builds from 14.3.0-canary.77
• Also affects: React Router, Waku, RedwoodSDK, Parcel, Vite RSC plugin

Immediate actions:

• Upgrade React to 19.0.3, 19.1.4, or 19.2.3 immediately
• Update Next.js to 16.0.7, 15.5.7, 15.4.8, 15.3.6, 15.2.6, 15.1.9, or 15.0.5
• Monitor for unusual multipart/form-data POST requests consistent with Next.js Server Actions / RSC endpoints
• Check logs for E{"digest" error patterns indicating exploitation attempts
• Review server processes for unexpected Node.js child processes

Exposure: ~310,500 Next.js instances on Shodan (US, India, Germany, Japan, Australia)

CVE-2025-20393 | Cisco Secure Email Gateway

Risk Score: 99 (Very Critical) | Active exploitation by UAT-9686

Why this matters: Chinese threat actors are actively compromising email security infrastructure to establish persistent access and pivot into internal networks.

Affected products: Cisco Secure Email Gateway and Secure Email and Web Manager running AsyncOS

Immediate actions:

• Apply Cisco's security updates immediately
• Monitor Spam Quarantine web interface access logs
• Check for modifications to /data/web/euq_webui/htdocs/index.py
• Hunt for AquaShell, AquaPurge, and AquaTunnel indicators
• Review outbound connections to suspicious IPs

Known C2 infrastructure: 172.233.67.176, 172.237.29.147, 38.54.56.95 (inactive)

Key Takeaways

• Intelligence drives better decisions. High-performing teams use threat intelligence not just for detection, but to inform strategic business decisions and communicate risk to leadership.
• Maturity means efficiency. Advanced programs focus on automation, high-fidelity indicators, and cross-functional collaboration—freeing analysts to concentrate on strategic initiatives.
• Information overload is the top challenge. Teams need better integrations and AI-powered tools to transform massive data volumes into actionable insights.
• AI will reshape the analyst role. While junior analysts won't be replaced, their workflows will evolve significantly as AI augments their capabilities.

Recorded Future recently hosted two webinars to unpack key insights from the 2025 State of Threat Intelligence Report and hear directly from customers who are putting these findings into practice.

Based on survey responses from 615 cybersecurity executives and practitioners, the report showed clear industry trends. Threat intelligence spending is up, with 76% of organizations spending over $250,000 annually and 91% planning to increase spending in 2026. Even more critically, 87% said they expect to advance the maturity of their threat intelligence programs over the next two years.

But what does maturity actually look like in practice? Our customers offered candid perspectives on how they're turning intelligence into impact.

Intelligence as a strategic asset

Our webinar panelists noted that the availability of rich threat intelligence has transformed how their organizations approach decision-making. According to Jack Watson, Senior Threat Intelligence Analyst at Global Payments, “Understanding that one alert opened and one alert closed does not necessarily equate to one single adversary being stopped” has led his team to take “a much more holistic approach to looking at problems.”

Omkar Nimbalkar, Senior Manager of Cyber Threat Research and Intelligence at Adobe, said, “Once you start doing this work day in and day out, you uncover patterns in your environment. You uncover what your posture looks like, where your true risk resides, and you can use that as a means to inform the business on the changing threat landscape for better decision-making.”

Ryan Boyero, Recorded Future’s Senior Customer Success Manager, said context and storytelling are key benefits of threat intelligence. “You can have a precursor or malicious activity that has occurred,” he said, “but without threat intelligence, you can’t really tell the story or paint the picture to deliver to senior leadership in order to help make the best and informed decisions possible.”

How threat intelligence delivers organization-wide value

Nimbalkar said his team provides tailored threat intelligence to business units and product teams across Adobe so they can monitor for specific behavioral activities and block specific threats in their environments.

Boyero shared that Recorded Future customers in EMEA use threat intelligence to educate leadership. “We're able to inform leaders,” he said. “We're able to speak with executives, get them in the room, not so much scare them that a situation could happen or has happened, but ultimately just educate and let them know that this is what Recorded Future is able to do and how we can bring success to the table.”

Erich Harbowy, Security Intelligence Engineer at Superhuman, said that in addition to educating leaders about risk, his team also uses threat intelligence to show the value of their work. “Not only am I using this very current news, I am also using the statistics that come along with that,” he said. “How much damage occurred during the first attack that was similar to this? And are [my adversaries] done? Are they coming back?”

Harbowy appreciates Recorded Future for providing those insights for postmortems and follow-ups with executives. “How do I prove my worth?” he said. “Give me the intel.”

The anatomy of a mature threat intelligence program

According to Nimbalkar, maturity comes when the foundational tactical and operational work is complete. He said that advancing a threat intelligence program is all about efficiency and optimization, including making sure you have high-fidelity indicators so your noise-to-signal ratio is reduced and you have higher-quality detections, understanding who your adversaries are and how they’re targeting you, getting in front of stakeholders and engaging with cross-functional teams, and collecting metrics on everything you do.

“Once you have figured out all these workflows, automated as much as you can, optimized and made it efficient, and then you focus more on risk reduction across the environment and more on strategic initiatives, that’s a very good maturation,” he said.

Jack Watson of Global Payments described threat intelligence maturity as the ability to ingest and action intelligence. “It’s never been easier to ingest data, but it’s also never been harder to sift through [that data]. So we’re seeing more mature organizations developing automated workflows, developing custom capabilities to do collection and action, and using AI in unique ways.”

Pathways to advancing maturity

Nick Rainho, Senior Intelligence Consultant at Recorded Future, said that the key to advancing maturity is having solid intelligence requirements. “Especially if you’re working with limited resources, go for the low-hanging fruit and ensure that the intelligence you’re pulling in is relevant to senior leadership’s priorities.”

Ryan Boyero agreed that maturity success is predicated on understanding leadership’s key requirements. “And then, how are we able to work towards that greater good and define success together?”

Top challenges for CTI teams

The panelists agreed that information overload is a critical challenge for today’s CTI teams. “More data is better than less,” said Watson, “but you have to be able to whittle it down or it’s useless.”

Nimbalkar said that with new tools in the market, advancements in AI, and the exponential growth in the volume of data, teams need vendors that can provide better integration to make data more actionable. And Rainho agreed, calling for better out-of-the-box integrations between intelligence tools so security teams can consume intelligence in the location and manner that works best for them.

Looking to the future of threat intelligence

When asked how they think the threat landscape will evolve and how technology will evolve with it, the panelists shared a number of predictions. They believe AI will enable CTI teams to fight AI-powered threats at scale. Third-party risk management will become an even more critical discipline for proactive defense. Digital threats will continue to outpace physical threats. And while junior analysts won’t be replaced by AI, their jobs will look very different as they use AI to augment their workflows.

Watch the recordings of the North America and EMEA webinar sessions to learn more, and download the 2025 State of Threat Intelligence Report to see how your peers are evaluating, investing in, and operationalizing threat intelligence.

Key Takeaways

• Declining payments, evolving tactics: Ransomware groups made less money in 2025 despite a 47% increase in publicly reported attacks, pushing them to adopt new approaches to extract payment, namely, DDoS-as-a-Service offerings, insider recruitment, and gig worker exploitation.
• Insider threats are rising: With stolen credentials, vulnerability exploitation, and phishing still dominating initial access, ransomware operators are increasingly turning to native English speakers to recruit corporate insiders—a trend likely to accelerate if layoffs continue into 2026.
• Global expansion underway: Recorded Future predicts 2026 will mark the first year that new ransomware actors operating outside Russia outnumber those within it, reflecting the rapid globalization of the ransomware ecosystem.

The ransomware paradox: More attacks, less money

By most accounts, ransomware groups made less money in 2025 than in 2024, both in overall payments and average payment size. This occurred despite a significant increase in attack volume: according to Recorded Future Intelligence, publicly reported attacks rose to 7,200 in 2025 compared to 4,900 in 2024, demonstrating a 47% increase.

For context, Recorded Future classifies both encryption attacks and data theft attacks with an extortion component under the ransomware umbrella. While exact numbers are difficult to isolate, approximately 50% of all attacks we track fall into the data theft and extortion category.

This declining profitability is driving ransomware groups to expand and evolve their tactics. Here are three trends organizations should prepare for heading into 2026.

Trend 1: DDoS services return to the RaaS model

With affiliates earning less and many ransomware operators abandoning the Ransomware-as-a-Service (RaaS) model to operate independently, remaining RaaS operations must offer more value to attract and retain affiliates. One increasingly common differentiator: bundled DDoS services.

The newly formed Chaos ransomware group (distinct from the older group of the same name) exemplifies this trend, providing DDoS capabilities to all affiliates. While this tactic isn't new—for example, REvil previously offered similar services—it fell out of favor for a period. Now, with fewer ransom payments to share, RaaS operators are reintroducing premium services to maintain their affiliate networks.

• What this means for defenders: Organizations should ensure their DDoS mitigation strategies account for attacks that may accompany ransomware incidents. The pressure tactics are becoming multi-pronged.

Trend 2: Insider recruitment attempts are accelerating

Stolen credentials, vulnerability exploitation, and phishing remain by far the most common initial access vectors for ransomware groups, with social engineering as a distant but growing fourth method. However, there has been a notable increase in ransomware groups working with native English speakers to recruit corporate insiders.

The most public example came earlier this year when a ransomware group attempted to recruit a reporter at the BBC. But this represents only the visible tip of a larger trend. Private reporting indicates that insider recruitment attempts increased significantly throughout 2025 and will likely continue growing, especially if workforce reductions at major companies persist into 2026.

• What this means for defenders: Insider threat programs should be evaluated and strengthened. Employee awareness training should address the possibility of external recruitment attempts, and organizations should monitor for anomalous access patterns that could indicate insider-facilitated attacks.

Trend 3: Gig workers as unwitting attack vectors

According to a recent FBI advisory, ransomware groups have begun exploiting gig work platforms to carry out attacks when remote methods fail. In one documented case, an attacker successfully executed a social engineering help desk scam but couldn't install their tools remotely due to security controls. Their solution: recruiting a gig worker through a legitimate platform to physically enter corporate offices and steal data.

The gig worker was unaware they were working for hackers, believing they were performing a legitimate IT task. The targeted employee thought they were assisting someone from the help desk. While this attack vector remains rare, the accessibility and global reach of gig work platforms means other groups could replicate this approach with minimal effort.

• What this means for defenders: Physical security protocols should account for social engineering scenarios involving legitimate-looking third parties. Verification procedures for on-site IT work deserve renewed scrutiny.

Looking ahead: One big prediction for 2026

The ransomware ecosystem has seen tremendous growth among actors and groups operating outside of Russia.

Recorded Future believes that 2026 will be the first year the number of new ransomware actors outside Russia exceeds those emerging within it. This doesn't indicate a decline in Russian-based operations; instead, it reflects how dramatically the global ransomware ecosystem has expanded.

The bottom line: Strengthen your ransomware defenses

Understanding emerging ransomware tactics is the first step toward defending against them. To stay ahead of threat actors and protect your organization:

• Explore Recorded Future's Ransomware Mitigation Solution for end-to-end visibility into your ransomware exposure across the attack lifecycle.
• Read our latest Insikt Group® research on ransomware trends, threat actor TTPs, and emerging attack vectors.
• Download the Proactive Ransomware Mitigation eBook for actionable strategies to identify, investigate, and prioritize cyber threats.

Hacktivist group Anna’s Archive claims to have scraped almost all of Spotify’s catalog and is now seeding it via BitTorrent, effectively turning a streaming platform into a roughly 300 TB pirate “preservation archive.”

On its blog, the group states:

“A while ago, we discovered a way to scrape Spotify at scale. We saw a role for us here to build a music archive primarily aimed at preservation.”

Spotify insists that the hacktivists obtained no user data. Still, the incident highlights how large‑scale scraping, digital rights management (DRM) circumvention, and weak abuse controls can turn major content platforms into high‑value targets.

Anna’s Archive claims it obtained metadata for around 256 million tracks and audio files for roughly 86 million songs, totaling close to 300 TB. Reportedly, this represents about 99.9% of Spotify’s catalog and roughly 99.6% of all streams.

Spotify says it has “identified and disabled the nefarious user accounts that engaged in unlawful scraping” and implemented new safeguards.

From a security perspective, this incident is a textbook example of how scraping can escalate beyond “just metadata” into industrial‑scale content theft. By combining public APIs, token abuse, rate‑limit evasion, and DRM bypass techniques, attackers can extract protected content at scale. If you can create or compromise enough accounts and make them appear legitimate, you can chip away at content protections over time.

The “Spotify scrape” will likely be framed as a copyright story. But from a security angle, it serves as a reminder: if a platform exposes content or metadata at scale, someone will eventually automate access to it, weaponize it, and redistribute it.

And hiding behind violations of terms and conditions—which have never stopped criminals—is not effective security control.

How does this affect you?

There is currently no indication that passwords, payment details, or private playlists were exposed. This incident is purely about content and metadata, not user databases. That said, scammers may still claim otherwise. Be cautious of messages alleging your account data was compromised and asking for your login details.

Some general Spotify security tips, to be on the safe side:

• If you have reused your Spotify password elsewhere or shared your credentials, consider changing your password for peace of mind.
• Regularly review active sessions on streaming services and revoke anything you do not recognize. Spotify does not offer per-device session management, but you can sign out of all devices via Account > Settings and privacy on the Spotify website.
• Avoid unofficial downloaders, converters, or “Spotify mods” that ask for your login or broad OAuth permissions. These tools often rely on the same kind of scraping infrastructure—or worse, function as credential-stealing malware.

---

We don’t just report on threats – we help protect your social media

Cybersecurity risks should never spread beyond a headline. Protect your social media accounts by using Malwarebytes Identity Theft Protection.

Key Takeaways

• Digital threats now originate far beyond the perimeter. Identity exposure, brand impersonation, and attacker coordination across the open, deep, and dark webs create risks that traditional tools cannot detect early enough.
• Context is the foundation of effective detection. Raw alerts and isolated indicators offer little clarity. Real-time intelligence turns noise into actionable insight.
• Modern digital threat detection (DTD) requires visibility across the external digital environment. The earliest warning signs of ransomware, credential theft, and phishing campaigns appear long before internal alerts fire.
• Analysts need automation to keep pace. High alert volumes and false positives overwhelm SOC teams. Automated enrichment, correlation, and prioritization significantly reduce investigation time and alert fatigue.
• Recorded Future operationalizes intelligence at enterprise scale. The Intelligence GraphⓇ, Digital Risk Protection, and deep SIEM/SOAR/EDR integrations deliver immediate context, organization-specific visibility, and unified detections, improving time-to-detect, time-to-contain, and overall resilience.

Why Digital Threat Detection Requires a New Approach

Today’s cyber threats evolve too quickly and appear across too many digital touchpoints for isolated tools or static detection rules to keep up. SOC teams must contend with:

• High alert volumes from SIEM, EDR, cloud telemetry, identity systems, and external sources.
• Evolving adversary techniques, including automated attacks and infrastructure that changes by the hour.
• Expanding attack surfaces driven by SaaS adoption, third-party dependencies, social platforms, and cloud-native architectures.
• Alert fatigue from manually sifting through noise to find high-risk signals.

As a result, organizations often struggle to distinguish meaningful threats from the constant noise of daily security events.

Digital threat detection (DTD) addresses this challenge by shifting focus from isolated internal signals to continuous identification, analysis, and prioritization of threats across an organization’s entire digital ecosystem. Unlike traditional perimeter-focused detection, which relies on firewalls, antivirus, and static rules, DTD recognizes that modern threats originate from external infrastructure, supply chains, cloud environments, identities, brand assets, and the open web.

The shift from reactive, point-in-time monitoring toward a proactive, intelligence-led model gives defenders the context they need to understand not just what is happening, but why it’s happening and what to do next. This article will serve as a comprehensive guide for security professionals, defining DTD and exploring the essential tools, methodologies, and practices required to build a proactive and intelligent security program.

Understanding the Modern Digital Threat Landscape

To build an effective digital threat detection program, security teams must understand where modern threats originate and how attackers operate.

Key Threat Vectors Beyond the Perimeter

Leaked credentials and account takeover attempts (stolen identities)

Compromised identities are now the most common entry point for attackers. Credentials harvested from stealer logs, breach dumps, or phishing toolkits often circulate online long before defenders know they’re exposed.

Brand impersonation, domain spoofing, and phishing campaigns

Attackers increasingly weaponize an organization’s public presence and create look-alike domains, fraudulent social profiles, or cloned websites to exploit user trust. These impersonation campaigns often serve as the launchpad for credential harvesting, malware delivery, and social engineering operations.

Vulnerability exploitation and zero-day threats in the external attack surface

Public-facing assets such as web applications, cloud workloads, exposed services, and third-party integrations are constantly probed for misconfigurations and unpatched vulnerabilities.

Dark web chatter and early warning signs of planned ransomware or DDoS attacks

Long before a ransomware deployment or DDoS attack hits production systems, signals often surface in underground communities. Threat actors discuss tools, trade access, or signal interest in specific industries and regions.

Why an Intelligence-Driven Approach is Better

For years, security programs centered their detection efforts on internal activity: log anomalies, endpoint alerts, authentication failures, and other signals that only appear after an attacker is already inside the environment. This approach is inherently reactive. It reveals what is happening within your systems, but not what is forming outside your walls or who may be preparing to target you next.

Digital threat detection reverses that model. Instead of waiting for internal symptoms of compromise, it looks outward at the behaviors and infrastructure, and intent of adversaries operating across the broader digital ecosystem. This expanded perspective allows teams to identify threats earlier in the kill chain, sometimes before any malicious activity reaches corporate networks.

The real advantage comes from context. Raw data on its own is ambiguous: an IP address, a file hash, a domain registration. With intelligence layered on top, those fragments become meaningful. Context exposes intent, and intent enables defenders to prioritize, escalate, or respond with precision rather than guesswork.

Essential Digital Threat Detection Tools and Technologies

Modern digital threat detection depends on a collection of tools that work together to surface early warning signals and provide the context you need to validate threats quickly.

Threat Intelligence Platforms: The Engines of Context

No human team can manually aggregate, cross-reference, and analyze the amount of threat data emerging across the web every minute. A modern threat intelligence platform automates this work, transforming massive volumes of raw, unstructured information into intelligence that analysts can act on immediately.

Threat intelligence platforms collect data from a wide range of external sources and standardize it into a usable format. Sources include:

• Open web reporting
• Underground forums
• Dark web marketplaces
• Malware sandboxes
• Threat feeds
• Researcher data

Once the data is normalized, the platform enriches it with context, such as:

• Relationships between indicators
• Associations with known threat actors
• Infrastructure reuse
• Activity targeting specific industries or regions

This enrichment process turns isolated artifacts into a coherent picture of adversary behavior, revealing intent, relevance, and potential impact in ways raw data alone cannot.

Security Orchestration, Automation, and Response (SOAR)

While threat intelligence provides the context needed to understand potential risks, SOAR platforms help teams take action on that intelligence quickly and consistently. These tools automate routine tasks that would otherwise consume analyst time, ensuring that high-priority threats receive attention without delay.

Key SOAR capabilities include:

• Enriching alerts with additional context from internal systems (SIEM, EDR, IAM, cloud telemetry)
• Blocking malicious indicators across firewalls, endpoints, cloud environments, and identity systems
• Initiating takedown workflows for harmful domains or impersonation infrastructure
• Coordinating actions across multiple security tools to ensure a unified response
• Documenting each step of the investigation for reporting and compliance

By automating the mechanics of response, SOAR platforms allow analysts to focus on higher-value decision making rather than repetitive execution, reducing dwell time and improving overall response efficiency.

Endpoint Detection and Response (EDR) & Security Information and Event Management (SIEM) Integration

EDR and SIEM platforms provide the internal vantage point of a digital threat detection program.

EDR monitors activity directly on endpoints, capturing details such as running processes, file modifications, and other behaviors that may indicate compromise on individual devices. SIEM systems, by contrast, collect and correlate logs from across the entire environment, including authentication systems, cloud services, applications, and network devices.

Together, these tools create a continuous stream of telemetry that reveals what is happening inside the organization, from process activity and login events to cloud logs and network traffic. When this internal data is correlated with intelligence about adversary infrastructure, active campaigns, or malicious tooling observed in the wild, EDR and SIEM can separate routine activity from signs of actual threats.

Modern platforms increasingly apply AI and machine learning to enhance this capability. Instead of relying solely on static signatures or predefined rules, they learn normal behavior across users and systems and identify subtle deviations that signal compromise.

Overcoming the Analyst’s Biggest Pain Points

Today’s threat landscape places enormous pressure on analysts. Internal alerts arrive faster than they can investigate them, and the earliest indicators of an attack often originate in places no traditional tool monitors.

The Drain of Alert Fatigue and False Positives

High alert volumes are a major driver of analyst burnout. Much of the day is spent triaging notifications with little context, forcing analysts to manually determine which events represent real threats and which are routine activity. The repetitive, high-stakes nature of this work is exhausting and increases the likelihood that critical signals will be missed.

The only reliable way to cut through this noise is to improve the quality of context surrounding each alert. When telemetry is paired with intelligence that explains adversary intent, infrastructure, and behavior, analysts can immediately see which signals matter and which can be safely deprioritized.

The Blind Spots of External Risk

Much of the activity that signals an impending attack happens beyond the reach of traditional security monitoring. Early warning signs often surface on the deep and dark webs, in criminal marketplaces, inside closed forums, and across fast-moving social platforms.

These external environments are frequently where the most actionable signals appear first. Credential dumps, access sales, discussions about targeting specific industries, and the creation of malicious infrastructure often occur long before any internal alert fires. Without insight into this external ecosystem, organizations are effectively blind to the earliest stages of an attack. And monitoring these spaces manually is nearly impossible at scale.

Recorded Future: Operationalizing Digital Threat Intelligence at Scale

Recorded Future’s approach to digital threat detection delivers real-time intelligence at enterprise scale, closing the visibility gaps that make modern detection so difficult and giving you the context you need, the moment you need it.

Real-Time Context from the Intelligence GraphⓇ

The Intelligence GraphⓇ addresses the fragmentation of global threat data, one of the most persistent challenges in modern security operations. Threat activity unfolds across millions of sources, including:

• Open web
• Dark web marketplaces
• Malware repositories
• Technical feeds
• Network telemetry
• Closed underground forums

No analyst team could manually track, interpret, and connect this information at the speed attackers operate. The Intelligence GraphⓇ solves this problem by continuously indexing and analyzing this vast ecosystem in real time. It structures billions of data points into clear relationships among threat actors, infrastructure, malware families, vulnerabilities, and targeted industries. Because these connections are made automatically, the platform can deliver immediate, decision-ready context on any indicator.

Comprehensive Digital Risk Protection for External Threats

Real-time context helps analysts understand what a threat is and who is behind it. But detection isn’t only about interpreting indicators; it's also about discovering specific threats against your organization across the broader internet.

Recorded Future’s Digital Risk Protection (DRP) solution focuses on the same external spaces where global threat activity occurs, but applies a different lens: it monitors those environments for anything tied to your brand, domains, executives, or employees. This targeted approach ensures you see early signals of impersonation, credential theft, or emerging attacks long before they reach your internal systems.

Accelerating Time-to-Action through Integrated Intelligence

Recorded Future accelerates detection and response by delivering high-fidelity intelligence directly into the tools analysts already rely on.

An extensive ecosystem of pre-built integrations and flexible APIs connect directly with every major SIEM, SOAR, and EDR platform. These integrations feed enriched threat context, dynamic Risk Scores, and prioritized intelligence into the tools analysts already use.

Collective InsightsⓇ adds a layer of visibility that other tools cannot provide. It consolidates detections from across your SIEM, EDR, SOAR, IAM, and other security platforms into a single view, then enriches them with high-fidelity Recorded Future intelligence.

This approach connects internal alerts to one another and exposes relationships that would remain hidden when each tool operates in isolation. By identifying MITRE ATT&CK® tactics, techniques and procedures (TTPs) and attributing malware, it surfaces attack patterns you can only see from an aggregated view.

Smarter, Faster Security Decisions

Recorded Future delivers the automated, contextual intelligence needed to identify risks the moment they emerge and empower teams to respond with confidence.

By unifying internal telemetry with real-time global threat insight and organization-specific targeting data, the platform enables smarter prioritization, faster action, and dramatically less noise.

These intelligence-driven workflows directly improve core detection metrics such as time-to-detect (TTD) and time-to-contain (TTC), giving organizations a measurable way to demonstrate progress and strengthen operational resilience.

Strengthen your security program and move toward intelligence-driven operations with confidence. Explore how Recorded Future can support your Digital Threat Detection strategy.

Introduction

The cybersecurity landscape is rapidly growing in scale and complexity. Enterprises face a rising tide of sophisticated threats that cannot be contained by traditional, reactive defenses alone. With AI and automation lowering the barrier to entry for attackers exploiting new avenues, there is more opportunity than ever for disruptive, high-volume attacks.

The need for organizations to mature their threat intelligence capabilities is clear, but the road to get there isn’t always easy. Recorded Future’s 2025 State of Threat Intelligence Report found that only 49% of enterprises currently consider their threat intelligence maturity as advanced, yet 87% expect to make significant progress in the next two years.

This gap between today’s capabilities and tomorrow’s ambitions reflects a familiar challenge: organizations have plenty of threat data, but struggle to connect, automate, and operationalize it effectively across teams and tools.

Based on insights from the report, here is what enterprises can expect when it comes to threat intelligence in 2026.

Key Trends Driving Threat Intelligence Evolution

There are several key trends set to shape threat intelligence in the coming year, and organizations wanting to prioritize maturity should be on the lookout for partners that embrace and evolve with these currents in mind.

• Vendor Consolidation for Unified Intelligence: Enterprises are looking to reduce tool fragmentation by consolidating threat intelligence vendors and feeds into a single platform. A unified approach promises a “single source of truth,” making it easier to operationalize intelligence across the organization.
• Deeper Integration into Security Workflows: Organizations want threat intelligence deeply embedded in their existing security stack rather than as a siloed feed. In fact, 25% of enterprises plan to integrate threat intelligence with additional workflows (e.g. IAM, fraud, GRC) in the next two years to broaden their reach.
• Automation and AI Augmentation: To cope with accelerating threats and volumes of data, teams are embracing automation in threat intelligence. The future lies in machine-speed analysis that automatically correlates and enriches intelligence so analysts can focus on high-level judgment.
• Fusion of Internal and External Data: Over a third of organizations (36%) plan to combine external threat intelligence with data from their own environment to gain better insight into risk posture (and even benchmark against peers).

Challenges Holding Team Backs Today

Despite this forward momentum, many enterprise teams still struggle with persistent challenges that hinder their threat intelligence efforts.

• Integration Gaps: Fragmented ecosystems remain a top concern. Nearly half of organizations (48%) cite poor integration with existing security tools among their biggest pain points.
• Credibility and Trust Issues: Data means little if analysts don’t trust the intelligence. Half of enterprises say verifying the credibility and accuracy of threat intelligence is a major challenge.
• Signal-to-Noise Overload: With huge volumes of alerts and feeds, 46% of enterprises struggle to filter relevant insight from noise. This information overload hampers visibility into real threats, drains team efficiency, and contributes to analyst burnout.
• Lack of Context for Action: Even when threat data is available, 46% of organizations lack the context needed to translate it into meaningful risk insights or actionable priorities.

These barriers help explain why many programs plateau at an intermediate maturity. Teams may ingest more data sources over time, but still fall short on the automation, integration, and context needed for truly advanced, predictive intelligence.

Envisioning Threat Intelligence in 2026: Proactive, Integrated, and Business-Aligned

In the near future, leading enterprises will treat threat intelligence not as a side task but as a strategic function integrated into business processes. This means embedding threat insights directly into risk assessments, vulnerability management, and even board-level decisions on security (notably, 58% of organizations already use threat intelligence to guide business risk assessment decisions today).

Instead of simply reacting to incidents after they occur, advanced threat intelligence programs will analyze patterns and emerging trends to warn of potential attacks before they fully materialize. This doesn’t mean magically “knowing the future,” but sharpening awareness by connecting subtle signals across many sources and mapping them to one’s environment.

Human analysts will still be central for this kind of work, though their capabilities will be augmented by AI such that detection and response happen at machine speed. Intelligence platforms will automatically enrich new indicators, correlate them with ongoing events, and even trigger protective actions in real time—all with analysts overseeing the entire process.

Ultimately, a mature program in 2026 will be measured by the outcomes it enables and the risk it reduces for the organization. This means protecting the assets, uptime, and reputation the business cares about, and improving decision quality at all levels of management.

Implications for 2026 Security Budgets and Investments

As threat intelligence becomes more central to security strategy, it’s also becoming a bigger line item in budgets. In fact, 91% of organizations plan to increase their threat intelligence spending in 2026, reflecting its critical role in an era of escalating cyber threats.

One likely area for these increased funds is platform consolidation. Many teams are reevaluating their myriad point solutions and considering a move to more integrated platforms that unify multiple sources and use cases, reducing complexity and cost over time.

Another likely investment will be in automation and AI capabilities. With cyber talent scarce and alert volumes ever-increasing, it will be vital to budget for tools that automate threat intelligence workflows end-to-end. From data collection and enrichment to triage and even initial response, automation will be key to doing more with the same team.

Organizations should also ensure that new investments deliver contextual intelligence tailored to their business. It’s not enough to simply buy more feeds or tools that spit out data; the value lies in solutions that fuse internal data with external threat feeds and apply analytics to highlight what matters most.

That said, not every organization will have the same needs and challenges. The key to fully maximizing ROI will be aligning spending with the organization’s biggest gaps and pain points. If credibility of data is a major challenge, invest in sources with proven reliability or validation features. If integration is a key issue, focus spending on consolidation projects or appropriate vendor services.

Security teams should also establish clear metrics (such as reduced incident response time or incidents prevented) to measure the impact of threat intelligence investments. For example, over half (54%) of organizations now measure success by improved detection and response times, making it a top metric for demonstrating value delivered by threat intelligence initiatives.

Charting the Course to 2026

Enterprise threat intelligence is undoubtedly maturing and becoming more ingrained in security programs, yet much work still remains. Nearly half of organizations may call themselves “advanced” today, but truly predictive, integrated intelligence at scale is still a goalpost ahead. In looking toward 2026, security leaders should double down on the fundamentals that drive intelligence maturity: integration, automation, and alignment with business priorities.

By breaking down silos between tools and teams, trusting and acting on intelligence through improved data credibility and context, and continually measuring what works, teams can evolve from reactive defense to an anticipatory, intelligence-driven security posture.

So what are some practical next steps? First, it’s wise to benchmark your organization’s current program to identify gaps and opportunities. Tools like Recorded Future’s Threat Intelligence Maturity Assessment provide a structured way to evaluate where you stand today and get tailored recommendations on how to improve.

With that insight, you can develop a roadmap that includes the right people, process, and technology investments to operationalize threat intelligence in the most efficient way. Keep the big picture in mind: the ultimate aim is to see more threats, identify them faster, and take action to reduce risk before damage is done. With a thoughtful strategy and an eye towards these trends, organizations can chart a course from today’s challenges to a more proactive and resilient threat intelligence function in 2026 and beyond.

November 2025 saw a significant 69% decrease in high-impact vulnerabilities, with Recorded Future's Insikt Group® identifying 10 vulnerabilities requiring immediate attention, down from 32 in October.

What security teams need to know:

• Fortinet leads concerns: Two critical FortiWeb vulnerabilities (CVE-2025-64446 and CVE-2025-58034) are under active exploitation
• LANDFALL spyware campaign: Threat actors weaponized Samsung's image processing flaw (CVE-2025-21042) for zero-click Android attacks
• Public exploits proliferate: Seven of ten vulnerabilities have public proof-of-concept code available
• OS Command Injection and Out-of-bounds Write were tied as the most common weakness types

Bottom line: The reduced volume shouldn't signal reduced vigilance. November's vulnerabilities demonstrate that threat actors favored quality over quantity in their exploitation campaigns.

Quick Reference: November 2025 Vulnerability Table

All 10 vulnerabilities below were actively exploited in November 2025.

Table 1: List of vulnerabilities that were actively exploited in November based on Recorded Future data (Source: Recorded Future)

Key Trends: November 2025

Vendors Most Affected

• Fortinet dominated with two critical FortiWeb vulnerabilities, both enabling remote exploitation
• Microsoft faced a kernel-level race condition affecting all modern Windows versions
• Samsung saw the weaponization of an image processing vulnerability for sophisticated mobile attacks
• Additional affected vendors: Gladinet, Google, Oracle, WatchGuard, CentOS, and Autonomy (OpenPLC)

Most Common Weakness Types

• CWE-78 – OS Command Injection (tied for first)
• CWE-787 – Out-of-bounds Write (tied for first)
• CWE-284 – Improper Access Control
• CWE-362 – Race Condition
• CWE-306 – Missing Authentication for Critical Function

Threat Actor Activity

LANDFALL Android spyware campaign marked November's most sophisticated operation:

• Exploited CVE-2025-21042 for zero-click remote code execution on Samsung devices
• Targeted Middle Eastern countries (Iraq, Iran, Turkey, Morocco) with commercial-grade spyware
• Deployed via weaponized DNG image files through WhatsApp
• Achieved persistent device compromise without user interaction
• Demonstrated advanced anti-analysis and SELinux bypass capabilities

Priority Alert: Active Exploitation

These vulnerabilities demand immediate attention due to confirmed exploitation in the wild.

CVE-2025-64446 | Fortinet FortiWeb

Risk Score: 99 (Very Critical) | CISA KEV: Added November 14, 2025

Why this matters: Unauthenticated attackers can bypass authentication entirely and create administrative accounts. With 4,768 exposed FortiWeb instances globally, this represents a critical internet-facing risk.

Affected versions: FortiWeb 8.0.0-8.0.1, 7.6.0-7.6.4, 7.4.0-7.4.9, 7.2.0-7.2.11, 7.0.0-7.0.11

Immediate actions:

• Apply Fortinet's security updates (8.0.2, 7.6.5, 7.4.10, 7.2.12, or 7.0.12)
• Monitor for POST requests to /api/v2.0/cmd/system/admin%3F/../../../cgi-bin/fwbcgi
• Check for unauthorized admin accounts created since October 2025
• Review logs for Base64-encoded CGIINFO headers
• Disable HTTP/HTTPS on internet-facing interfaces if patching is delayed

Exposure: ~4,768 FortiWeb instances visible on Shodan (Netherlands, US, Germany, Italy, Peru)

Key Takeaways

• Static vendor checks fall short: Traditional, point-in-time third-party risk management practices (e.g. annual questionnaires) leave organizations blind to emerging vendor threats between audits. Continuous monitoring is now a must.
• Five common risk scenarios: Supply chain attacks, widespread software vulnerabilities, hidden fourth-party dependencies, vendor credential theft, and vendor instability each illustrate how “trusting” vendors can lead to breaches or business disruptions.
• Intelligence-driven defense: Recorded Future’s platform provides real-time visibility into your vendor ecosystem—from dark web credential leaks to fourth-party relationships—enabling proactive mitigation before incidents impact your organization.
• From trust to verification: The solution is to move from static trust to continuous verification. By continuously assessing vendors’ cyber and business health (and even integrating intelligence into workflows like ServiceNow), security leaders can vastly strengthen their vendor risk management framework.

Your Vendor Ecosystem Is a Black Box: It’s Time to Turn on the Lights

For CISOs and risk leaders, the attack surface now goes far beyond the footprint of the business. It’s a sprawling web of SaaS vendors, software suppliers, MSPs, payment processors, logistics partners, and niche fourth parties your vendors rely on. Every connection expands risk—often outside direct visibility. In other words, your security may only be as strong as your weakest vendor or partner.

Traditional third-party risk management (TPRM)—static security questionnaires and annual audits—cannot keep pace. They describe what a vendor claimed their security looked like months ago, not what it is right now. Meanwhile, the most damaging events (supply chain attacks, zero-day exploitation, credential resale, concentration failures) unfold in hours and days, not quarters.

This gap between point-in-time paperwork and real-time risk is why third-party exposure has become a primary vector for catastrophic breaches and business outages.

This article will highlight and analyze 5 real-world third-party risk examples. For each, we'll show why traditional methods fail and how continuous, real-time third-party risk management and threat intelligence is the only effective prevention.

5 Third-Party Risk Examples and How to Prevent Them

Modern vendor risk comes in many forms. Let’s explore five common scenarios—and how proactive measures can stop them:

Type 1: The Software Supply Chain Attack

The Scenario: One of the most damaging third-party risks is a software supply chain attack. This occurs when threat actors breach a trusted software vendor’s development environment and secretly inject malicious code into a legitimate, digitally signed software update. The tainted update, a “Trojan horse,” is then distributed to the vendor’s customers, giving the attacker access into thousands of networks at once.

Real-World Example: The SolarWinds Orion breach is a quintessential case. In 2020, nation-state hackers compromised SolarWinds’ build pipeline and inserted malware into an Orion software update. The malicious update, being validly signed, was pushed to around 18,000 customers, including numerous government agencies and Fortune 500 companies, who all gladly installed it, thereby granting the attackers insider access to their systems.

Why Traditional Methods Fail: A standard vendor security questionnaire or audit would never have caught this. SolarWinds had passed assessments and appeared reputable. The update itself was digitally signed and appeared “trusted” to antivirus scanners and other controls. In short, you cannot audit your way out of a risk that’s been inserted into a trusted product’s software supply chain.

The Intelligence-Led Solution: Preventing a supply chain attack means detecting subtle warning signs before the breach fully unfolds. Recorded Future’s platform continuously monitors for early indicators tied to your vendors. If threat actors known for targeting CI/CD pipelines start discussing or probing one of your software vendors, you’d know. If intelligence suggests a vendor’s code-signing certificate may be compromised, you’d get an alert. Armed with this foresight, you could elevate that vendor’s risk status, scrutinize their software updates more closely, and even hunt for indicators of compromise in your environment before the breach becomes public knowledge.

Type 2: The Widespread Third-Party Vulnerability

The Scenario: A critical software vulnerability (often a zero-day) is discovered in a common component that many of your vendors use. It could be an open-source library, a popular IT tool, or a cloud service. You have no direct visibility that your suppliers rely on this component. Attackers quickly develop an exploit and start compromising organizations at scale via this flaw, long before most victims even realize they’re exposed through their third parties.

Real-World Example: The MOVEit Transfer zero-day (exploited by the Cl0p ransomware group) and the Log4j “Log4Shell” vulnerability are perfect examples of this risk. In the case of MOVEit, a single bug in a widely used file-transfer product led to the mass theft of data from thousands of companies, many of whom weren’t even direct customers of MOVEit, but their vendors were. Similarly, the Log4j flaw impacted countless businesses indirectly because software used by their contractors and providers included the vulnerable library.

Why Traditional Methods Fail: This is fundamentally a technology visibility problem. A point-in-time survey asking your vendors “Do you use MOVEit?” is too little, too late. By the time you send out a questionnaire and get a reply (if you get one at all), attackers may have already exploited the vulnerability and exfiltrated data. No organization can manually track every piece of software in their extended vendor ecosystem through periodic check-ins. In the MOVEit incident, many companies had no idea they were at risk until news of data breaches surfaced. Traditional vendor risk management simply isn’t designed to monitor technical exposure in real time.

The Intelligence-Led Solution: Defending against widespread vulnerabilities requires connecting two dots instantly: what’s vulnerable and who in your supply chain is using it. This is where an intelligence platform shines. Recorded Future’s approach combines technical attack surface intelligence with real-time vulnerability tracking. It continuously scans the internet to map out the external-facing tech stack of your third parties. The moment a new critical vulnerability is disclosed, Recorded Future’s intelligence automatically checks which of your vendors are running that technology. You receive an immediate, prioritized alert such as: “CRITICAL: 15 of your third-party vendors are exposing servers running [the vulnerable software]. Prompt them to apply patches or mitigations immediately.”

Type 3: The Fourth-Party & Concentration Risk

The Scenario: Sometimes the biggest risk in your vendor ecosystem isn’t with your direct third parties, but with their key dependencies. A “fourth party” is a vendor of your vendor, and if one that many of your critical vendors rely on goes down, it can create a single point of failure. A single outage can cascade up the chain, disrupting operations even when direct vendors appear secure.

Real-World Example: The 2021 ransomware attack on Kaseya’s VSA remote monitoring and management platform is a textbook case. Kaseya primarily served managed service providers (MSPs), who in turn delivered IT services to thousands of downstream customers. When attackers exploited Kaseya VSA, they were effectively able to push ransomware out through those MSPs to many organizations that had no direct relationship with Kaseya at all—they only “knew” their MSP. A single fourth-party dependency became the pivot point for a broad, multi-industry disruption.

Why Traditional Methods Fail: If you looked at each of your primary (third-party) vendors in isolation, they all might have passed your security reviews with flying colors. What the traditional assessment missed was that ten of those vendors all relied on the same subcontractor for a critical function, a critical audit blind spot. Most organizations only discovered their exposure to Kaseya after MSP-delivered systems were already encrypted. Without continuous visibility into your vendors’ vendors, this kind of concentration risk remains invisible until it’s too late.

The Intelligence-Led Solution: The only way to manage fourth-party and concentration risk is through continuous mapping of your vendors’ vendors, coupled with dynamic risk scoring. Recorded Future’s Third-Party Intelligence solution automatically identifies and maps these Nth-party relationships throughout your supply chain. In practice, this means if a critical fourth-party suffers a breach, you won’t be finding out via the news days later. Instead, your intelligence dashboard would immediately show that entity’s risk score spiking from, say, a modest 50 to a critical 99. This timely insight gives you a head start to activate business continuity and incident response plans. You immediately know exactly which of your vendors are impacted and can work to contain the fallout.

Type 4: The Vendor Credential Compromise

The Scenario: Not all third-party attacks involve sophisticated malware or supply chain tampering. Sometimes hackers just log in through the front door. In this scenario, a threat actor steals valid credentials from one of your vendors and uses those to access your systems. Perhaps an employee at a smaller, “low-risk” vendor, like an HVAC contractor, falls victim to a phishing email or unknowingly runs info-stealer malware on their laptop. Their VPN login or application credentials to your network get quietly harvested and sold on the dark web. An attacker buys the login, bypasses your multi-factor authentication, and walks into your network posing as a legitimate third-party user.

Real-World Example: This tactic was at the heart of the high-profile 2023 breaches of MGM Resorts and Caesars Entertainment, where attackers initially gained access via a third-party IT support vendor’s compromised VPN credentials.

Why Traditional Methods Fail: A vendor security questionnaire cannot prevent an individual at a partner company from clicking a phishing link or using a weak password. Your vendor might have all the right policies on paper, but those policies are irrelevant the moment an attacker has a valid username and password in hand. Traditional TPRM programs are about vetting a vendor’s security controls and compliance, but they don’t provide real-time awareness of things like a password leak or dark web sale of access related to that vendor.

The Intelligence-Led Solution: The key to stopping a credential-based breach is catching those compromised credentials before they are used against you. This calls for continuous identity-centric intelligence. Recorded Future’s Third-Party Intelligence module includes automated monitoring of a wide range of sources, from dark web forums to infostealer logs and criminal marketplaces, specifically watching for any mention of your organization’s partners and their accounts. The moment a set of credentials associated with one of your vendors appears in an illicit context, you receive a high-priority alert. Your team can immediately revoke or reset that vendor account and investigate the extent of access. This is the definition of proactive defense: you’re effectively shutting the door on the attacker before they can walk through it.

Type 5: The Operational & Financial Instability Risk

The Scenario: Sometimes the greatest third-party risk is a vendor’s operational or financial collapse. Consider a scenario where a critical vendor suddenly encounters a non-cyber crisis like bankruptcy, a major lawsuit or regulatory sanction, a natural disaster, or even a geopolitical event that halts their business. From your security team’s perspective everything looked fine, but virtually overnight this partner’s failure threatens to grind your business to a halt.

Real-World Example: A headline-grabbing case occurred with the sudden collapse of Silicon Valley Bank (SVB) in March 2023. SVB wasn’t attacked by hackers; it suffered a bank run and shut down in a matter of days. Companies that used SVB as a banking partner or for credit found themselves unable to access funds or process payroll, creating a cascade of operational and financial issues.

Why Traditional Methods Fail: A standard security questionnaire or compliance-focused vendor review is utterly blind to this category of risk. Your CISO’s third-party risk process likely doesn’t include reviewing a vendor’s financial statements or monitoring news about their executives’ legal troubles—nor should it, in a traditional model, since those are outside the classic IT security scope. As a result, organizations were caught off-guard by SVB’s collapse. A vendor that looked perfectly green from a security control standpoint turned out to be a huge business continuity threat. This kind of event exposes an “edge case” risk that isn’t an edge case at all: vendors can introduce strategic and financial risks that security teams and vendor managers often aren’t tracking.

The Intelligence-Led Solution: Truly comprehensive third-party risk management means monitoring all-source intelligence on your vendors, not just cyber indicators. Recorded Future’s Third-Party Intelligence platform is built to ingest and analyze a broad spectrum of data about companies. This includes real-time monitoring of global news media, credit ratings and financial filings, changes in executive leadership, legal filings, sanctions lists, regulatory watchlists, and more. By defining “risk” holistically, the platform can alert you to significant non-cyber events that may impact your vendors. These signals give your security, risk, and procurement teams time to react, whether that means activating contingency plans, finding alternate suppliers, or engaging leadership to address the issue.

The Solution: Move from “Trust” to “Continuous Verification”

The five examples share a theme: “trust” is not a control. Vendor attestations and annual audits don’t capture rapidly changing third-party conditions—exploits, credentials, dependencies, and financial shocks. To answer why third-party risk management is important: it’s no longer a “vendor” problem. It’s your attack surface, your data, and your reputation on the line.

This is why security leaders are shifting from a trust-but-verify model to a model of continuous verification, replacing blind trust with live intelligence.

Moving to continuous verification means supplementing or replacing periodic vendor check-ins with real-time intelligence and automation. This is where Recorded Future’s approach comes in. Recorded Future acts as a “risk radar” that’s always on, giving you a 360-degree, real-time view of your third-party ecosystem. It uniquely integrates multiple intelligence streams—threat intelligence, attack surface intelligence, and third-party risk intelligence—into one platform.

• Know which CVEs matter today across your ecosystem with Vulnerability Intelligence and exploit-in-the-wild context.
• Detect compromised vendor access with Identity Intelligence and automated revocation workflows.
• Map fourth-party dependencies and track concentration with Third-Party Intelligence risk scoring.
• Operationalize all of this via integrations to SIEM/SOAR/EDR and GRC/TPRM workflows (e.g., ServiceNow) so that risk evidence triggers action.

Recorded Future is the only platform connecting disparate, live third-party intelligence into a single, real-time view that answers the question:

“Which of my vendors poses the greatest risk to my business—right now?”

Ready to replace point-in-time vendor questionnaires with continuous verification? Schedule a personalized demo, and our experts will show you how the Recorded Future platform provides a complete, real-time picture of your vendor ecosystem.

Last updated on 9 December.

A critical vulnerability in React Server Components is allegedly being actively exploited by multiple Chinese threat actors, Recorded Future recommends organizations patch their systems immediately.

What's Happening

CVE-2025-55182, dubbed "React2Shell," affects React Server Components versions 19.0, 19.1.0, 19.1.1, and 19.2.0 in several Meta packages. Amazon's AWS Threat Intelligence team reported on December 4 that Chinese threat groups including Earth Lamia, Jackpot Panda, and several untracked clusters are actively exploiting this vulnerability. However, AWS has not provided any further evidence for these attributions beyond IP addresses allegedly used by these threat groups. At this stage, Insikt Group cannot exclude the possibility that the same threat group might still be using the IP address 206[.]237[.]3[.]150, but we are currently unable to verify AWS’s attribution to Earth Lamia.

The vulnerability stems from unsafe payload deserialization at React Server Function endpoints. When successfully exploited, attackers can execute arbitrary code through crafted HTTP requests, potentially leading to complete backend compromise.

There are now multiple publicly available exploit scripts (I forked one on GitHub here) for the React and Next.js vulnerabilities (CVE-2025-55182 and CVE-2025-66478).

The underlying issue is data serialization/deserialization, which evoked thoughts about a blog I wrote in 2016, addressing the same issue (at the time, the topic was CVE-2015-4852, a serialization flaw in Java objects that affected Oracle and Apache products).

2 Risk Takeaways

• The exploit pattern repeats because serialization is a straightforward method for transferring data, and developers typically use what works. Coders use different languages and frameworks, yet the same class of vulnerability persists. The upstream opportunity here is for universities to aggressively drive security into all programming courses.
• Everyone is a coder now, and security domain expertise has never been more important. Every business function will include AI-assisted coders, supercharging productivity and efficiency. LLMs don’t need to stop for human input, but understanding internet plumbing, tools, platforms, and security implications is now crucial. The most valuable employees can use AI for 10x+ impact AND catch potential issues as humans become the AI-copilots.

Technical Causation

• Serialization is seductive: It’s the easy path for passing complex objects across trust boundaries (client ↔ server, service ↔ service). Developers reach for it because it “just works” (until it catastrophically doesn’t).
• Framework abstraction hides the danger: Some percentage of Next.js developers using Server Actions are unaware that they’re invoking a custom serialization protocol. They’re calling a function. The risk is invisible until it’s exploited.
• The ecosystem never learns collectively: Java shops learned (painfully) about gadget chains and ObjectInputStream. However, that institutional knowledge didn’t necessarily transfer to Node.js/React developers building RSC implementations a decade later.

The Threat

The attack surface has expanded once again. In 2015, we were tracking exploit chatter on Chinese forums appearing weeks before CVEs. Now, there are double-digit public GitHub repositories with weaponized exploit code within days of disclosure. Agentic workflows will soon compress that window down to minutes. The time-to-exploitation will amount to the time defenders spend reading about any new high-severity vulnerability.

Defender Considerations for CVE-2025-55182 / CVE-2025-66478

• Attackers differentiate vulnerable App Router targets from safe Pages Router sites by checking for window.__next_f vs __NEXT_DATA__. Your asset inventory should already know which flavor you’re running.
• The vulnerability lives in the Flight protocol deserialization. If you’re not using Server Actions, consider disabling them. If you are, the endpoint (Next-Action header targets) is where to focus WAF rules.
• Hunt for anomalous POST requests with Next-Action headers containing malicious multipart payloads targeting __proto__ or unusual serialized JSON structures. The exploit exfils via base64 in error digests.
• The core issue is in react-server-dom-webpack, react-server-dom-parcel, and react-server-dom-turbopack. Custom RSC implementations outside Next.js are equally exposed.
• RCE means immediate credential harvesting from environment variables, lateral movement via cloud metadata endpoints, and persistence via scheduled tasks or cron jobs. IR playbooks should assume full compromise.

For “Vibe Coders”: A Conversation with Claude Opus 4.5

Q: What’s a proper input validation methodology for deserializing data to avoid security failures?

A: Native serialization formats (Java ObjectInputStream, PHP unserialize, Python pickle, .NET BinaryFormatter) reconstruct objects with behavior, which means arbitrary code can execute during deserialization. The fix is using formats that only transfer data, not executable structures.

Safe alternatives:

1. Data-only interchange formats

• JSON (most common)
• Protocol Buffers (Google)
• FlatBuffers (zero-copy, fast)
• MessagePack (binary JSON, compact)
• CBOR (binary, IETF standard)

These parse into primitive types and structures—no object instantiation, no code execution.

2. Schema validation on top

• JSON Schema
• Zod / Yup (JavaScript)
• Pydantic (Python)
• marshmallow (Python)

Define exactly what shape the data must have. Reject anything that doesn’t match before your application logic ever sees it.

3. When you need objects, build them explicitly

Instead of:

obj = pickle.load(untrusted_data) # dangerous

Do:

data = json.loads(untrusted_data) # safe parse
obj = MyClass(**validated_schema(data)) # explicit construction

4. For config/markup

• TOML instead of YAML (no code execution tags)
• JSON5 for human-readable config
• Avoid YAML’s !!python/object or Ruby’s !ruby/object—or use safe loaders exclusively (yaml.safe_load())

The one-liner: If your serialization format can reconstruct arbitrary types, it’s a vulnerability. Stick to dumb data formats + explicit object construction.

Q: Do modern LLMs know how to securely validate data by default?
A: Technically, yes, practically, it’s complicated.

What’s in the training data:

• Security research, CVE databases, OWASP guidance
• Countless blog posts about pickle/unserialize dangers
• Secure coding documentation from every major vendor

So if you ask directly—”Is pickle safe for untrusted data?”—any decent LLM will say no.

The problem:

LLMs are also trained on millions of Stack Overflow answers, tutorials, and GitHub repos that use pickle.load(), unserialize(), and yaml.load() because they’re convenient. When someone asks, “How do I save this object to a file and load it later?” the pattern-match often lands on the dangerous-but-common approach.

What I’ve observed:

• Ask for “the fastest way” → often gets insecure code
• Ask for “production-ready” or “secure” → better results
• Ask without security context → coin flip

The real gap:

LLMs don’t reason about security—they predict likely tokens based on training. They won’t spontaneously say “wait, where is this data coming from?” the way an experienced developer might. The threat model isn’t automatic.

Practical advice for developers using LLM assistants:

• Explicitly ask for secure alternatives
• Ask “What are the security implications of this code?”
• Treat generated code like untrusted input—review it

So yes, the knowledge is in there. But so is a decade of insecure examples, and the LLM doesn’t inherently prioritize one over the other unless prompted.

Stay Ahead with Recorded Future

Learn how Recorded Future's Vulnerability Intelligence tracks CVE disclosures, monitors exploit availability across GitHub and underground forums, and prioritizes vulnerabilities based on active threats to your environment. The platform compresses your response window from days to hours.

In the wake of the Salesforce/Gainsight breach (kudos to Salesforce for transparently sharing indicators of compromise and updated progress on remediation), third-party cyber and exposure risk is top of mind for many CISOs. Professional services firms are often overlooked in this context, with disastrous consequences.

Law firms, specifically, are particularly vulnerable to creating downstream risk impacts given the nature and purpose of legal services, and adversary targeting is on the rise.

The Industrial Consolidation of Legal Sector Attacks

The numbers paint a stark reality. Twenty % of US law firms were targeted by cyberattacks in the past year, with 56% of breached firms losing sensitive client information. The average breach cost reached $5.08 million, representing a 10% year-over-year increase that excludes long-term reputational damage and client defection.

RansomHub has emerged as 2025’s dominant threat after absorbing talent from disrupted groups like LockBit and ALPHV/BlackCat. By offering affiliates a 90/10 profit split versus the standard 70/30, they’ve attracted the most capable operators in the underground economy. Qilin’s Rust-based ransomware has specifically targeted legal entities with encryption-resistant payloads, making recovery nearly impossible.

The chart below, derived from Recorded Future analyst notes tracking ransomware extortion sites, illustrates the growth in ransomware targeting by industry, with legal firms remaining the number one target.

These aren’t opportunistic attacks. Threat actors now maintain “dwell times” exceeding weeks inside firm networks, systematically identifying crown jewel intelligence before triggering extortion events. Industrialization means attackers understand exactly what creates maximum leverage: M&A intelligence during active deals, litigation strategies before trial, and decades of retained client data across multiple matters.

Recorded Future telemetry from the past quarter indicates that over 20 observed legal or legally adjacent firms have malware communicating with malicious command-and-control (C2) servers. While the observed traffic was 24 hours or less for some firms, other organizations saw persistence above 5 days. Certainly, a malicious implant does not equate to a full breach and exfiltration of client-sensitive data; however, it is a valuable signal to monitor for changes in third-party and fourth-party risk.

Infographic depicting recent malware dwell times in global legal firm victims

When Privilege Becomes Your Adversary’s Weapon

Courts have systematically eroded attorney-client privilege protection for breach investigations, creating a dangerous trap where forensic reports become ammunition for adversaries. The Capital One decision ordered production of Mandiant’s forensic report because the investigator served “business purposes” rather than pure legal advice.

The cascade accelerates through “sword and shield” waiver doctrine. Any use of breach investigation findings, even citing them in discovery responses, can trigger a subject matter waiver, requiring disclosure of all privileged communications related to threat assessment and remediation strategy. The 2024 Samsung Data Breach ruling made this explicit: sharing reports with 15 executives indicated business decision-making use, defeating privilege.

Federal Rule of Evidence 502 creates additional exposure when companies share incident reports with regulators. The 2023 Covington & Burling case saw the SEC subpoena the firm for names of 298 publicly-traded clients whose data “may have been exfiltrated,” though a court eventually ruled that only seven clients had to be named, it did establish that law firms cannot completely shield client identity from regulators, and those clients could then face SEC investigation for failure to disclose their counsel was breached.

M&A Intelligence Monetization at Scale

When Berkeley Research Group was hit by ransomware in March 2025 during a $700 million leveraged buyout by TowerBrook Capital Partners, the attack exposed M&A intelligence across hundreds of concurrent deals. This wasn’t just data theft; it was a systematic opportunity for market manipulation.

Academic research quantifies the damage. The Intralinks/Cass Business School study found 8-10% of M&A deals leak annually, with leaked deals achieving 47% median premiums versus 27% for non-leaked deals, which is a 20 percentage point difference worth millions per transaction. Only 49% of leaked deals complete versus 72% of non-leaked deals.

The Tyler Loudon case (2024) demonstrated the benefits of access when the defendant stole M&A information from his attorney wife, resulting in insider trading charges.

The Systematic Failure to Assess Professional Services Risk

Only 30% of law firms report clients asking them to complete security questionnaires (not that attestations are a wholly competent method for determining exposure risk), compared to a near-universal requirement for SaaS vendors. This exemption culture may stem from relationship bias and the misconception that “they’re not a tech vendor” despite law firms operating technology-intensive businesses.

The data concentration goes untracked. A single firm may hold M&A details, employee PII, trade secrets, litigation strategies, regulatory issues, and executive compensation across multiple business units that operate independently. The Orrick breach (2023) exposed 637,000+ individuals precisely because the firm aggregated data from employment litigation, mergers and acquisitions (M&A) transactions, and patent filings.

Retention amnesia compounds the risk. Lawyers traditionally “keep everything forever” due to a risk-averse culture, and potential regulatory requirements. Data from cases in the 1990s may still exist on unpatched legacy servers. Each year of retention adds cumulative breach exposure, yet enterprises rarely ask law firms about deletion policies or data locations.

Strategic Actions for Enterprise Defense

Treating professional services firms as high-risk technology vendors requires structural changes to vendor management frameworks.

• Eliminate standing exemptions: Subject law and consulting firms to the same security requirements as SaaS vendors, including SOC 2 verification, independent audits, and quarterly assessments, without granting relationship-based waivers.
• Map concentration risk: Identify all professional services vendors with data access across business units. Calculate total organizational exposure when single firms hold aggregated intelligence across HR, legal, finance, and compliance matters.
• Audit fourth-party dependencies: Require disclosure of critical vendors, including MSPs, cloud providers, SaaS vendors, and document management systems. A breach of fourth-party infrastructure becomes your breach through the use of API tokens, credential harvesting, and VPN pivoting.
• Establish time-bound access: Implement purpose-limited credentials that expire at the conclusion of a matter. Eliminate long-lived access that persists in engagement reports and consulting code repositories.
• Define retention requirements: Specify data deletion periods in contracts with confirmation requirements. Audit compliance quarterly, as many firms retain data indefinitely on legacy systems.
• Deploy breach detection: Place honeytokens in systems accessible to professional services firms. Establish 24-48 hour notification SLAs with emergency credential rotation capabilities.
• Create specialized incident response protocols: Develop playbooks specifically for law firm breaches addressing privilege complications, litigation exposure assessment, and regulatory notification requirements.
• Use threat intelligence to map services firms’ domain and IP space. Use the infrastructure map to monitor and alert on observed traffic between malware implants and command-and-control (C2) infrastructure. Recorded Future's Third-Party Intelligence automates this monitoring across your entire vendor ecosystem, providing real-time alerts when professional services firms show compromise indicators. Combined with Ransomware Mitigation capabilities, organizations can track ransomware group TTPs, monitor extortion sites, and receive early warnings when vendors appear on leak sites. Immediately notify affected service providers, disable organizational access, and assist in remediation.

Wrap-Up

The evidence from 2025 makes the stakes undeniable. With 21 law firm breaches in just the first five months of 2024 and incidents like Williams & Connolly’s nation-state compromise and Berkeley Research Group’s ransomware during active M&A, the pattern is clear.

When your law firm holding decades of critical data gets breached, you don’t have a vendor incident. You have a strategic intelligence compromise with multi-year competitive implications that traditional third-party risk frameworks didn’t adequately contemplate, as they exempt “trusted advisors” from the security scrutiny their data concentration demands. The shift from relationship-based trust to risk-based verification isn’t optional; it’s survival.

Learn how Recorded Future's Ransomware Mitigation and Third-Party Intelligence solutions work together to protect against cascading vendor risk. From tracking ransomware groups targeting legal firms to monitoring your vendors for real-time compromise indicators, you can detect and respond to vendor compromises before they cascade into your organization.