SAN FRANCISCO — Artificial intelligence company Anthropic has suspended access to its newest advanced AI models after receiving a U.S. government directive tied to national security concerns, the company said on Friday, marking one of the most significant restrictions yet imposed on access to frontier AI systems.
Anthropic said it had been ordered by U.S. authorities to halt access to its Fable 5 and Mythos 5 models for foreign nationals. According to the company, the scope of the directive made it impossible to selectively enforce the restriction, leading it to disable the models for all users while it seeks to comply with the order. Access to the company's other AI models remains unaffected, Anthropic said.
In a public statement, Anthropic said the U.S. government had issued an export-control directive citing national security authorities. “The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance,” the company said.
The company said it had not been provided detailed information about the alleged security risks. Anthropic stated that it understood government officials were concerned that safeguards on one of the models could potentially be bypassed, allowing it to assist in identifying software vulnerabilities. The company disputed the severity of those concerns and said the evidence presented to it was limited.
The suspension came only days after the release of Fable 5, one of Anthropic’s latest AI offerings. Mythos 5, a more advanced system with significant cybersecurity-related capabilities, had already been subject to tighter controls and scrutiny because of concerns about potential misuse.
Reports by multiple media outlets said the Commerce Department directive applies to foreign nationals both inside and outside the United States, including non-U.S. citizens employed by Anthropic. The order reflects growing government attention to the national security implications of highly capable AI systems and their potential use in cyber operations.
Anthropic said it supports government authority to intervene when deployments pose legitimate risks but argued that such actions should be transparent and grounded in technical evidence. The company described the current situation as a “misunderstanding” and said it hopes to restore access once the matter is resolved.
As of Saturday, the company said the two models remained offline while discussions with government officials continued. Details of any review process or timeline for restoring access remain unclear.
