Fable and Mythos are back. But U.S. AI policy is still a mess. | Fortune

The latest U.S. AI policy debate centers on Anthropic’s Fable and Mythos returning after export controls were rolled back, yet Fortune says the broader framework remains inconsistent. The article reports the government reversed course on Anthropic’s “Mythos” late Friday and lifted controls on “Fable” late on Tuesday, after the earlier restrictions had forced Anthropic to disable both models for all users. It notes the shift is likely welcomed by Anthropic and investors, and by defenders hoping to use Mythos for finding security flaws, but argues the wider policy still operates like a licensing regime despite official denial. Fortune says the U.S. is working with leading AI labs on explicit “voluntary standards” on cybersecurity to reduce the chance the government objects to public releases, and that Anthropic is also collaborating on a shared framework to assess risks from jailbreaks. The report adds that OpenAI is seeking a 5% stake in the company and mentions a reported breakthrough in compute efficiency, while Meta stock is rising on plans for a cloud business.