Why the US cyber agency is using Anthropic’s Mythos to audit government code

The US cyber defense agency is actively using Anthropic’s advanced artificial intelligence model, Mythos to audit government code for severe vulnerabilities- a move that shows signs of political backing for adopting the AI startup’s tools even as the company resolves its ongoing standoff with the White House.

According to sources, the Cybersecurity and Infrastructure Security Agency is using Mythos to scan government code repositories for flaws that could leave the door open for cybercriminals. At present, it has not been confirmed how much government code the agency has gone through the severity of the vulnerabilities it discovered.

Anthropic which has filed privately for a US initial public offering has had a volatile relationship with the US government. Relations reached a nadir in February after the San Francisco-based company refused to disable security protocols that prevented its AI from being used for autonomous weapons or domestic surveillance.

The National Security Agency is the US government’s espionage arm, which has been using Mythos despite the blacklist in April.

Claude Fable 5 is primarily a safeguarded version of Anthropic’s Claude Mythos engine-marking a major move for an AI program rivalling competitors OpenAI’s ChatGPT and Google’s Gemini.

Anthropic introduced a public version of Mythos called Fable to implement security controls after the White House demanded that it ban non-residents. Furthermore, this significant move triggered a temporary shutdown of the model, which was lifted last week.