Trump just found the worst way to regulate AI
TL;DR
Anthropic first shared Mythos only with vetted organizations and released Fable as a heavily restricted public version. The model reportedly beat earlier systems on benchmarks while refusing many cyber and biology requests. After Amazon warned officials about a possible jailbreak, the Trump administration imposed export controls. In practice, Anthropic had to take Fable offline because it could not reliably exclude foreign users or foreign-national employees.
Nauti's Take
This is not a credible blueprint for AI regulation; it is a warning sign. The US needs rules that measure dangerous capabilities and treat providers consistently, not a system where proximity to the White House becomes a safety criterion.
Anthropic is not automatically blameless just because it speaks the language of safety. But pulling a model from the market on opaque grounds is the worst mix of symbolism, economic self-harm, and discretionary power.
Briefingshow
The case separates serious AI safety policy from discretionary political control. If frontier models create new cyber or bio risks, governments need clear tests, independent review, and predictable thresholds. An executive takedown without transparent standards creates uncertainty for developers, customers, and international partners.
Tweets
Sources