CISA Uses Anthropic AI for Cybersecurity, Despite Tensions

The Cybersecurity and Infrastructure Security Agency, or CISA, is reportedly using Anthropic's advanced AI model, Mythos, to scan government software for vulnerabilities. This is happening even as the Pentagon and White House have had disagreements with Anthropic over how its AI should be used. CISA's audits aim to find bugs that could expose systems to foreign spies or cybercriminals. These scans have already uncovered a large number of vulnerabilities. What's interesting is the contradiction in US policy. While federal agencies like CISA are adopting Anthropic's AI for critical cybersecurity tasks, the Pentagon previously tried to blacklist Anthropic. This was because the company refused to allow its AI to be used for autonomous weapons or domestic surveillance. A judge later blocked this designation. The NSA has also been testing Mythos since April, with analysts reportedly impressed by its capabilities. The White House also demanded a temporary global shutdown of a public version of Mythos, called Fable, due to concerns about misuse. Access was restored last week. The bottom line is that despite these tensions, US agencies continue to rely on Anthropic's AI for important security work.