Anthropic abruptly disables Fable 5 and Mythos 5 after U.S. national security order

Anthropic has pulled access to its newest and most advanced AI models just days after launch, following a U.S. government directive tied to national security concerns.

The AI startup said Friday that it will “abruptly disable” access to Fable 5 and Mythos 5 for all foreign nationals after receiving an export control order from the U.S. government. The move affects users worldwide and comes at a sensitive moment for Anthropic, which confidentially filed for an IPO last month and has been positioning itself as one of the leading competitors to OpenAI.

“The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Anthropic models will not be affected,” Anthropic said in a statement.

According to Anthropic, government officials instructed the company to suspend access to the models without providing detailed evidence behind the national security concerns that triggered the order.

U.S. Orders Anthropic to Shut Down Its Most Advanced AI Models for Foreign Nationals

Anthropic said it believes the government’s concern centers on a possible method of bypassing, or “jailbreaking,” safeguards built into Fable 5. Those safeguards are intended to prevent the model from being used to identify software vulnerabilities that could be exploited in cyberattacks.

The company pushed back against the decision.

“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,” Anthropic said in a statement.

The dispute marks a sharp escalation in Washington’s approach to AI regulation. For years, U.S. export controls focused on restricting access to advanced semiconductors and computing hardware. This latest action targets access to AI models themselves, signaling a new front in the government’s effort to limit the spread of advanced AI capabilities.

A U.S. official confirmed that the Commerce Department issued an export control directive requiring Anthropic to suspend access to Fable 5 and Mythos 5 for foreign nationals.

The order arrives after months of tension between Anthropic and parts of the U.S. government.

Relations deteriorated earlier this year after Anthropic refused to allow its AI models to be used for domestic surveillance and fully autonomous weapons systems, according to the company. The government later placed Anthropic on a supply chain blacklist that is scheduled to take effect later this year.

Just days before the shutdown order, Anthropic unveiled Claude Fable 5, a model the company described as part of a new “Mythos-class” category of AI systems. Anthropic said the release included safeguards restricting use in areas such as cybersecurity. Some users criticized those restrictions as overly broad.

Security researchers have warned that highly capable AI systems could accelerate sophisticated cyberattacks if misused. Concerns are particularly acute in industries such as banking, where critical infrastructure often relies on complex software built over decades.

Anthropic said it worked with government agencies and outside experts before launching the models and argued that competing frontier AI systems demonstrate similar abilities in identifying minor software bugs.

“The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Anthropic models will not be affected,” the company said.

Anthropic described the situation as a misunderstanding and said it is actively working with government officials to restore access.

“If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers,” the company said.

The impact is already spreading beyond Anthropic itself.

Amazon Web Services said Friday that Anthropic requested the removal of access to the affected models for “all users in all regions.”

The order could create new questions for AI companies operating globally. Dean Ball, a former White House official who contributed to the administration’s AI Action Plan, said the directive appears broad enough to restrict access for virtually all non-U.S. citizens.

“This means you should expect to have to prove your citizenship to use Anthropic models,” Ball wrote on X.

The implications extend inside Anthropic’s own workforce. Several prominent members of the company’s leadership and research teams were born outside the United States, Reuters reported. The episode highlights a growing clash between AI developers and regulators over how much risk is acceptable before a model is released to the public.

Just two days before receiving the order, Anthropic publicly called for stronger oversight of advanced AI systems and supported government intervention for models that present unacceptable risks. Yet the company argues that the action taken against Fable 5 and Mythos 5 falls short of fair, evidence-based regulation.

Pentagon Chief Information Officer Kirsten Davies defended the government’s position.

“Some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation. America First. Always,” Davies wrote on X.

For Anthropic, the stakes extend far beyond two AI models. The company now finds itself at the center of a debate over who gets access to advanced AI, how governments should respond to emerging risks, and whether national security concerns can justify pulling a frontier AI system offline after it has already been released.