Anthropic Downplays AI Risks; White House Scorns Claims

Anthropic is downplaying security risks for its AI models, Mythos and Fable, after the White House banned their foreign use. This has drawn criticism from White House officials. Here's the thing: The White House implemented export controls on Mythos and Fable after Amazon's CEO reportedly informed the Trump administration that its researchers could bypass Fable's safety features, posing a national security risk. Anthropic then took both models offline entirely. What's interesting is Anthropic claims the flaw was a "narrow potential jailbreak" not unique to Fable and disagreed with the White House's actions. This response prompted scorn from Trump officials, who say Anthropic has previously raised fears about AI doomsday scenarios. One official stated the company seems "obsessed with safety for everyone except themselves." Another official said the Trump administration heard from nearly a half-dozen companies with concerns about Fable before taking action. The bottom line: This situation highlights ongoing tensions between AI developers and government regulators regarding AI safety and control.