US Blocks Anthropic's Most Advanced AI Models Over Security Fears

Anthropic abruptly shut down access to its two most capable AI systems on Friday after the U.S. government invoked export controls to block their use by foreign nationals, citing national security concerns.

The company disabled its Fable 5 and Mythos 5 models for all customers to comply with a directive from the Trump Administration. The order represents the first time the U.S. government has applied export controls directly to AI models rather than just the semiconductor chips that power them.

Why it matters

This unprecedented move signals a fundamental shift in how the U.S. government views advanced AI systems—not as commercial products but as strategic national security assets requiring the same export restrictions as weapons technology. The decision also exposes the vulnerability of countries dependent on American AI infrastructure, potentially accelerating global competition to develop independent AI capabilities.

The jailbreak concern

According to Anthropic, the government notified the company on Friday that it had become aware of a method to "jailbreak" Fable 5—a technique that could bypass the model's built-in safety guardrails. Anthropic said it received only verbal notice of what it characterized as a "potential narrow, non-universal jailbreak" and disputed that this warranted such drastic action.

The company had launched Fable 5 earlier in the week, describing it as reaching a new "Mythos-class" capability tier. At the time, Anthropic acknowledged the model's cybersecurity capabilities could "be misused to cause serious damage" without proper safeguards, particularly in identifying software vulnerabilities.

Commerce Secretary Howard Lutnick reportedly sent a letter to Anthropic CEO Dario Amodei formally notifying the company of the restrictions, according to The Wall Street Journal.

Escalating tensions with the Pentagon

The export control order comes amid an ongoing dispute between Anthropic and the Trump Administration. The conflict began when Anthropic refused to allow the U.S. military to use its AI models for fully autonomous weapons systems. In response, the Pentagon placed the company on a blacklist, deeming it too dangerous for government use.

Pentagon Chief Information Officer Kirsten Davies defended the decision on social media, stating that "some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation."

Global implications

The timing could prove particularly damaging for Anthropic, which is reportedly preparing for an initial public offering. The shutdown occurred the same day that SpaceX launched its IPO, achieving a $2.1 trillion market valuation.

Anton Leicht, a fellow at the Carnegie Endowment for International Peace, told TIME the move demonstrates how dominant U.S. AI capabilities have become. "The U.S. is so far ahead in the AI race already that it can afford to leave other countries behind as an afterthought of a domestic decision," he said.

The decision has sparked debate internationally about AI sovereignty—the ability of nations to control their own artificial intelligence infrastructure. British lawmaker Kanishka Narayan, minister for AI and Online Safety, called the ban a lesson in the importance of domestic AI investment.

However, Leicht argued that achieving AI independence remains unrealistic for most countries. "Only the U.S. builds frontier models, and the U.S. controls almost all the chips needed to train them," he noted, adding that even ambitious national AI projects might take over two years to reach the technological frontier.

Anthropic said it is working to restore access and believes the situation stems from a misunderstanding. Neither the Pentagon nor the Commerce Department responded to requests for comment.

These details were first reported by TIME.