The US government forced Anthropic to pull its two newest AI models, Fable 5 and Mythos 5, just three days after they launched.
The details:
- On June 12, citing national security, the government ordered Anthropic to cut both models off for everyone, including its own staff.
- Anthropic can't tell foreign from domestic users in real time, so both went dark worldwide. Its other models still work.
- The official reason was a reported "jailbreak" of the models' safety limits, which Anthropic says was narrow and already public.
- Reportedly, Amazon CEO Andy Jassy warned officials and helped spark it. Anthropic may not restore access for weeks, possibly months.
Why it matters: This is the world we are in now, where a lab can flip the switch on a tool you depend on overnight, with no warning and no say from you.
It will not be the last time.
The smart move is to not wire your whole workflow into one model. Stress-test new tools the moment they launch, and keep an open-source option you can run yourself as a backstop.