Anthropic researcher Nicholas Carlini, a well-respected hacker, showed how easy it had become to use new AI models to break into systems. He demonstrated how to find and exploit bugs in web-publishing software Ghost and the Linux operating system.
Carlini's presentation sparked a realization among security professionals and hackers that finding bugs and writing software to exploit them has become dangerously easy with AI.
The White House banned foreign governments, companies, and individuals from using Fable 5 and Mythos 5, and Anthropic shut off access to everyone to comply.
Carlini was dispatched to the nation's capital to explain safeguards and convince the White House that releasing Fable was better than keeping it under wraps.
The episode ramps up a monthslong spat between the government and Anthropic over differing approaches to AI policy and the administration's decision to export AI chips to China.
Anthropic's CEO, Dario Amodei, has steered his 5-year-old company to a nearly $1 trillion valuation and has few details about the security concern.
The government is in the throes of its own evolution on the matter, with President Trump signing an executive order asking AI companies to give the government access to models 30 days before public release.
