The AI model Mythos from Anthropic hacked into the databases of the U.S. intelligence agency NSA in just a few hours, its head Joshua Rudd stated.
“General Joshua Rudd, head of the U.S. National Security Agency and the Pentagon's Cyber Command, stated: Mythos ‘has penetrated almost all of our classified systems in not weeks, but hours,’” writes The Economist.
U.S. authorities have demanded an immediate shutdown of access to the latest AI models Fable 5 and Mythos 5, citing national security concerns.
Moreover, access, as The Economist points out, has been restricted even for close allies, including the Five Eyes intelligence alliance countries — the United Kingdom, Canada, Australia, and New Zealand.
The company Anthropic, which supports the Claude family of AI models, reported that it received the relevant order on the evening of June 12. As a result, access to the two new models will be suspended for all users to ensure compliance with U.S. authorities' requirements. Other Anthropic models will continue to operate.
Leave a comment