Anthropic, the AI safety company behind the Claude family of models, abruptly suspended worldwide access to two of its most advanced AI systems — Fable 5 and Mythos 5 — following pressure from the United States government tied to export control concerns. The shutdown happened over the weekend, according to ServeTheHome, which described the episode as "a wake-up call for the industry."
The action follows a broader push by the Trump administration to restrict who, globally, can use America's frontier AI tools. Al Jazeera reported that the US asked Anthropic to block international access to its top models. According to Newsweek, Trump's ban on foreign access to the new Anthropic models sparked a tech backlash.
The restrictions extend inside the US government as well. According to the Hindustan Times, President Trump on Friday ordered federal agencies — including the Pentagon, FEMA, US Citizenship and Immigration Services, and the Transportation Security Administration — to be barred from using certain AI tools connected to Anthropic.
Fable 5, described by The Probe as Anthropic's most powerful AI, was among the systems taken offline. That means both international users and American government agencies lost access to Anthropic's newest and most capable models at the same time.
Export controls have long governed physical goods like semiconductors and weapons systems, but applying them to AI software — essentially flipping a switch to revoke access for users worldwide — marks a significant and largely untested escalation in tech policy.
If governments can compel AI companies to instantly cut off access to frontier models, the assumption that powerful AI will remain globally and commercially available looks far shakier than the industry has counted on.