US Government Orders Anthropic to Shut Down Two Powerful AI Models Over Security Concerns

The U.S. government on Friday ordered Anthropic to immediately shut off access to two of its most powerful AI models — Claude Fable 5 and Claude Mythos 5 — citing national security concerns. Anthropic announced on X that it has complied, but it made clear it thinks the government is wrong.

The directive, received at 5:21 pm ET, forces the company to disable both models for all users worldwide, not just foreign nationals targeted by the export control order. Access to other Anthropic models is unaffected.

Mythos is Anthropic’s most capable model, previewed in early April and kept tightly restricted due to its exceptional ability to find security vulnerabilities in software. According to Anthropic, Mythos identified flaws in every major operating system and web browser it tested. Rather than release it broadly, the company launched Project Glasswing, sharing it with roughly 50 vetted organizations including Amazon, Apple, Google, Microsoft, and CrowdStrike for defensive cybersecurity work.

Fable 5, released just three days ago, was Anthropic’s answer to commercial pressure: a version of Mythos with guardrails blocking responses in high-risk areas like cybersecurity and biology, making it safe enough for general release. It immediately became the most capable publicly available AI model, according to Vals AI benchmarks.

The government framed the directive as an export control action, but in a blog post Anthropic says its understanding is that the underlying concern is a claimed jailbreak of Fable 5. So far, the company says, the government has provided only verbal evidence of a “potential narrow, non-universal jailbreak” — essentially prompting the model to read a specific codebase and identify software flaws. Anthropic adds that this capability level is already widely available in other public models, including OpenAI’s GPT-5.5, and is routinely used by cybersecurity professionals.

Anthropic also argues that its strongest safeguards operate through independent classifier systems separate from the model itself. Even if someone convinces Fable to keep talking past a refusal, the underlying protections against dangerous outputs remain in place. A review of recent usage found no evidence of those safeguards being successfully bypassed.

“We disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people,” the company wrote. “If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers.”

Anthropic is widely expected to pursue an IPO this year and has staked its public identity on being the safety-conscious alternative. The irony is not lost on observers: the very caution Anthropic displayed in restricting Mythos — promoted as too dangerous for public release — has attracted government scrutiny that could disrupt its business.

In April, OpenAI CEO Sam Altman told podcaster Ashlee Vance that Anthropic’s handling of Mythos amounted to “fear-based marketing.” “It is clearly incredible marketing to say, ‘We have built a bomb. We were about to drop it on your head. We will sell you a bomb shelter for $100 million,’” Altman said. He didn’t predict a government shutdown, but identified that when you spend months telling the world your AI is uniquely dangerous, the world tends to listen.