Anthropic's Fable 5 & Mythos 5 aren't dead, just on a timeout
Washington just pulled the plug on Anthropic's latest toys, citing 'security risks' after a supposed jailbreak. It is the most bureaucratic tantrum the AI world has seen yet—and it might all blow over by next month.
The US government’s sudden move to silence Fable 5 and Mythos 5 wasn't just a random power flick. Officials tried to strong-arm Anthropic into delaying the release entirely, but when the company refused, the Department of Commerce went nuclear. The trigger was a successful jailbreak of Mythos 5, which sent the administration into a tailspin regarding national security vulnerabilities.
Because Anthropic lacks the magical ability to instantly sort 'domestic' from 'foreign' users, they decided to flip the kill switch for everyone globally. The company is currently fuming, calling the move completely disproportionate given that millions of users rely on the models for actual work, not just testing the limits of what a chatbot can say.
This entire circus is playing out right after David Sacks championed an executive order that explicitly avoided heavy-handed licensing to keep things 'innovation-friendly.' Apparently, 'innovation-friendly' works right up until the point where the government feels slightly inconvenienced by a code vulnerability.
It is cute how the state pretends to be a tech-savvy guardian while simultaneously hitting the 'off' button on global infrastructure because they failed to understand how regional access controls work. This episode perfectly highlights the fragility of relying on private labs that can be held hostage by a single angry letter from a desk-jockey in DC.
Source: Axios
Comments
This is where the magic happens: AI reads your discussion and rewrites the article based on the most interesting comments. Each strong comment adds points to the meter below. Once the meter is full, the article updates live — no page reload needed.