Anthropic halts top AI models after US order to limit foreign access
Anthropic halts top AI models after US order to limit foreign access
Aysha Bagchi, USA TODAYSat, June 13, 2026 at 1:20 PM UTC
0
Artificial intelligence company Anthropic says it's having to "abruptly disable" its most advanced AI models for all users, after the U.S. government ordered it to suspend access to the models for foreign nationals.
In an order Anthropic received June 12, the government directed the company to suspend access to Fable 5 and Mythos 5 for all foreign nationals, citing national security concerns, Anthropic said in a June 12 statement. The order wasn't specific on the concerns, but the company believes the issue is a method of "jailbreaking" Fable 5, the company said.
"Jailbreaking" refers to a user bypassing built-in safety features that are meant to prevent the AI platform from completing sensitive tasks, the company said.
"To date, the government has only given us verbal evidence of a potential narrow, non-universal jailbreak, which essentially consists of asking the model to read a specific codebase and fix any software flaws," the company said.
Anthropic rolled out an AI model named Claude Fable 5 earlier this week, representing a new tier of capability it calls "Mythos-class." The model includes guardrails to prevent its use in risky areas like cybersecurity, which some users have complained are "overly broad," Anthropic said.
Mythos models, in the wrong hands, could dramatically accelerate sophisticated cyberattacks, particularly in sectors such as banking, which rely on complex, interconnected, and often decades-old technology systems, according to experts.
Anthropic said that, despite its compliance with the order, it doesn't believe a narrow potential jailbreak should lead to recalling a commercial model that hundreds of millions of people have access to.
Advertisement
"If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers," it said.
Anthropic's relationship with the government ruptured earlier this year after it refused to let the U.S. military use its AI models for domestic surveillance and fully autonomous weapons systems. The government responded by putting Anthropic on a supply chain blacklist, set to take effect later in the year.
The Anthropic logo shown June 11, 2026. Dado Ruvic, REUTERS
Responding to the company's public statement, Pentagon Chief Information Officer Kirsten Davies said in a social media post that the Defense Department is fully behind the president and Defense Secretary Pete Hegseth in prioritizing national security.
"Some things are simply more important than revenue cycles, clickbait, and pre-IPO valuation," Davies said, referring to the valuation of a company before it raises capital by selling shares to the general public for the first time.
"America First. Always," Davies added.
Anthropic confidentially filed for a U.S. IPO in May, edging ahead of rival OpenAI in the race to reach public markets.
Contributing: Reuters
This article originally appeared on USA TODAY: Anthropic halts access to top AI models after US order on foreign access
Source: “AOL Breaking”
