White House May Give Anthropic Mythos Access To Government Agencies, Report Says

Forbes has reached out to Anthropic and the Office of Management and Budget for comment.

What Did Jack Clark Say About Anthropic's Discussions With The Government?

Clark said Anthropic's stance is that the "government has to know about this stuff," seemingly referring to developments around AI and cybersecurity. Clark also commented on the dispute between Anthropic and the Department of Defense over the agency's decision to designate the company as a supply-chain risk. The designation was made after Anthropic declined to give the Pentagon unrestricted access to its AI models, resulting in contractors who work with the U.S. military being restricted from conducting any business with Anthropic. Clark said the matter between his company and the Department of Defense was a "narrow contracting dispute."

Key Background

Mythos has been provided to companies such as Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks. Anthropic has not committed to releasing the AI model to consumers, at least in the short-term, citing abilities bad actors could have by using Mythos' capabilities. Anthropic has claimed test runs of Mythos found some vulnerabilities in "every major operating system and web browser," autonomously identifying gaps in security completely unknown to software developers. The company has also said AI models have become so efficient at coding, they can outperform the most skilled humans at exploiting vulnerabilities in software. Michael O'Rourke, chief market strategist at JonesTrading, told Reuters if Mythos is as powerful as Anthropic says it is, "it just shows one, the weakness of the current software that's out there and two, that AI is still making incredible progress versus legacy software companies."