Anthropic pulls two AI models after U.S. export control order

Anthropic has taken its two newest AI models, Fable 5 and Mythos 5, offline following an export control order—a government regulation restricting technology transfer—issued by the Trump administration. The government sent a letter citing national security concerns on Friday, which was Friday evening in New York. The letter, which has not been made public, gave no specifics but required the company to ensure the models could not be used by any foreign nationals. Because Anthropic could not guarantee this restriction—noting that many of its own employees are foreign nationals—the company chose to pull both models entirely. Fable 5 had been more widely available to the public, while Mythos 5 was accessible to current Mythos users.

The sudden regulatory action follows reports that Amazon researchers allegedly found a way to bypass Fable 5’s guardrails, which are safety mechanisms implemented in AI models. Following these findings, Amazon CEO Andy Jassy raised concerns with the White House, triggering a rapid escalation over the weekend. The administration had previously labeled Anthropic a supply chain risk, and the two parties are currently involved in an ongoing lawsuit. Anthropic has pointed out that similar jailbreaks could also be found in several other AI models.

The regulatory action has drawn sharp criticism from industry observers and security professionals. According to TechCrunch podcast host Rebecca Bellan, cybersecurity experts “signed an open letter to ask Trump to revoke the order, and they say it’s actually dangerous to have to pull these advanced cybersecurity capabilities from network defenders in the U.S.” Bellan also noted that she feels the export control order might be retaliatory, suggesting that the White House is just looking out for any excuse to pummel Anthropic. Podcast host Sean O’Kane observed that Anthropic’s difficult relationship with the Trump administration stands apart from other AI laboratories. This distinct friction suggests that while Anthropic faces intense scrutiny, its competitors might not immediately face similar crackdowns.

Why it matters

The move highlights the growing friction between leading AI labs and the U.S. government, raising questions about whether export controls are being used as a tool for regulatory retaliation.