The Cybersecurity and Infrastructure Security Agency (CISA) has added three actively exploited vulnerabilities to its Known Exploited Vulnerabilities (KEV) catalog: flaws affecting Langflow, Apache Tomcat, and N-central. The KEV catalog exists specifically to flag security holes that attackers are already using in the wild, and this latest batch is no exception. Adding to the intrigue, researchers at Palo Alto Networks' Unit 42 have linked one of the exploitation campaigns to a hacking agent reportedly powered by DeepSeek, the Chinese AI model that has drawn scrutiny for its rapid adoption in both legitimate and malicious use cases.

For everyday internet users, KEV additions can feel like inside baseball. They're not. These entries directly affect the software running behind the websites, remote management tools, and AI platforms that businesses and service providers rely on to keep your data safe. When attackers exploit these systems before patches are applied, the fallout often lands on the people whose information those systems were supposed to protect.

Three New Entries in CISA's KEV Catalog

Each of the three flaws now flagged by CISA sits in a different corner of the software ecosystem, which is part of what makes this update notable. Langflow is an open-source platform used to build AI agent workflows, a category of software that has exploded in popularity as companies race to deploy AI tools internally. Apache Tomcat is a widely used web server and servlet container that underpins countless Java-based applications across enterprises and government agencies. N-central is a remote monitoring and management (RMM) platform used by IT service providers to administer client networks remotely, meaning a single compromised instance can potentially expose many downstream organizations at once.

By federal directive, agencies under CISA's Binding Operational Directive are required to patch KEV-listed vulnerabilities within a set window, reflecting how seriously the government treats confirmed active exploitation. Private organizations aren't bound by the same mandate, but security teams widely treat KEV listings as a strong signal to prioritize patching immediately rather than waiting for a routine update cycle.

This isn't the first time CISA has moved quickly to flag a flaw with real-world attackers already using it. The agency's Linux root access flaw disclosure from earlier this year followed a similar pattern: a vulnerability discovered, weaponized, and added to KEV in short order, underscoring how compressed the window between disclosure and exploitation has become across the software supply chain.

A DeepSeek-Powered Hacking Agent Enters the Picture

What sets this round of disclosures apart is Unit 42's finding that one of the exploitation campaigns appears to be driven by an AI hacking agent built on DeepSeek. AI-assisted attack tooling isn't new in concept, but a documented case tying a specific KEV-listed exploitation campaign to a named AI model is a meaningful data point. It suggests that the automation attackers use to find, weaponize, and deploy exploits against vulnerable systems like Langflow instances is maturing quickly.

The practical implication is speed. AI-assisted tooling can help attackers identify vulnerable, internet-facing instances of software like Langflow or Tomcat faster than manual reconnaissance ever could, shrinking the already narrow window defenders have to patch before exploitation begins.

Why Actively Exploited Flaws Matter for Privacy

The direct connection between infrastructure vulnerabilities and personal privacy is easy to overlook, but it's real. Tomcat servers and RMM platforms like N-central often sit at the heart of systems that store or transmit customer data, from login credentials to financial records. Langflow instances, meanwhile, are increasingly used to build AI applications that process sensitive user inputs, from customer service queries to internal business data. A successful remote code execution or unauthorized access exploit against any of these systems can give attackers a foothold to move laterally, exfiltrate data, or deploy ransomware, turning a technical vulnerability into a privacy incident affecting far more people than the original target organization.

What This Means For You

Most readers won't be personally patching a Tomcat server or an N-central deployment, but the ripple effects of these vulnerabilities can still reach you. If you use services built on affected platforms, whether that's a company's customer portal, an AI-powered support tool, or an IT provider managing your organization's network, a breach upstream can expose your data downstream. It's worth paying attention to breach notifications from services you use and treating any unexpected password reset requests or account activity alerts with extra caution in the weeks following news like this.

Actionable Takeaways

If you manage IT infrastructure, prioritize patching any Langflow, Tomcat, or N-central instances immediately, and check CISA's KEV catalog directly for the specific entries and remediation deadlines. If you're a general user, keep an eye on breach disclosures from companies you rely on, enable multi-factor authentication wherever it's offered, and treat this as a reminder that the software running quietly behind the scenes of your favorite services has a direct line to your personal data. Staying informed about KEV updates like this one is a small habit that pays off when the next actively exploited flaw makes headlines.