Anthropic AI Firm Suspends Fable 5 Access, US Government Cites National Security Concerns

Anthropic, the artificial intelligence firm, has suspended public availability of its Fable 5 AI tool and restricted access to its Mythos 5 iteration following intervention from the US government. This decision was prompted by Washington’s prohibition on any foreign national accessing the technology, citing national security concerns.

Executives from Anthropic are scheduled to meet with senior White House officials and the US Department of Commerce on Monday in Washington DC. The meeting is expected to include Dario Amodei, Anthropic’s chief executive, and Secretary Howard Lutnick, who leads the Commerce Department.

The company had previously acknowledged that Fable 5’s capabilities “exceed those of any model we’ve ever made generally available,” stating its public release “comes with risks.” Shortly after its release, the US government reportedly “became aware” of a potential “jailbreak” vulnerability, allowing the AI to be manipulated beyond its intended design. Anthropic, however, has only received “verbal evidence” of this purported flaw.

This latest dispute follows Anthropic’s recent legal challenge against the US Department of Defense regarding the usage parameters of its AI models. While a previous meeting with White House officials was characterised as “productive,” it remains uncertain if Fable 5 and Mythos 5 will be made accessible again after the upcoming discussions.