CISA Uses Anthropic's AI Mythos for Vulnerability Scanning Amidst US Policy Contradiction

The Cybersecurity and Infrastructure Security Agency (CISA) is reportedly deploying Anthropic’s advanced AI model Mythos to scan government software repositories for vulnerabilities, according to a report by Reuters.

The audits conducted by CISA’s Attack Surface Evaluation team, aim to recognise bugs that could expose systems to foreign spies or cybercriminals. The report further adds that the scans had already uncovered a large number of vulnerabilities, though details on severity remain undisclosed.

The revelation that CISA is actively using Mythos highlights the contradiction in US policy: while the Pentagon and White House spar with Anthropic over safeguards, federal agencies continue to embrace its AI for critical cybersecurity tasks.

Anthropic’s rocky ties with the government

The move comes as Anthropic navigates a tense relationship with the US government. Earlier this year, the Pentagon attempted to blacklist the company with a “supply-chain risk” designation after Anthropic refused to remove safeguards preventing its AI from being used for autonomous weapons or domestic surveillance. A judge later blocked the designation, easing tensions. Despite the standoff, US agencies have continued to adopt Anthropic’s tools. The NSA has been testing Mythos since April, with analysts reportedly impressed by its cybersecurity capabilities.

White House clash over public release

When Anthropic rolled out a public version of Mythos called Fable, the White House demanded that foreign users be banned, triggering a temporary global shutdown of the model. Access was restored only last week, underscoring the administration’s unease over the potential misuse of frontier AI systems.

Emails between Anthropic CEO Dario Amodei and Pentagon IT head

Recently, Court documents unsealed this week have pulled back the curtain on how one of America's hottest AI companies fell out with the Pentagon. The emails, first reported by The Wall Street Journal and released as part of Anthropic's ongoing litigation against the Department of Defense, trace a months-long exchange between Anthropic CEO Dario Amodei and Emil Michael, the undersecretary of defense for research and engineering. What began as a negotiation over how the military could use Claude ended with Anthropic labeled a "supply-chain risk," treatment usually reserved for foreign adversaries. The correspondence, spread across January and February 2026, tracks how a working relationship steadily came apart. On one side sat a safety-focused lab drawing hard lines. On the other, a Pentagon that wanted no strings attached. The filings don't just show who said what—they show where a $200 million relationship snapped. Amodei's position never moved. He wanted two hard limits on how the Department of War—the unofficial name the Pentagon has revived for the DoD—could deploy Anthropic's models: no fully autonomous weapons, and no domestic mass surveillance. In a January 15 email, he wrote that these boundaries set important guardrails against using powerful AI to spy on Americans, without blocking legitimate foreign intelligence work. The Pentagon wanted broader terms. It insisted on covering "all lawful uses," a phrase that leaves plenty of room.