Anthropic to disable its most advanced AI models after US order limiting foreign access | AI (artificial intelligence)
Anthropic said it would “abruptly disable” its most advanced AI models for all users after the US government suspended foreign nationals’ access to its models, citing national security concerns.
Anthropic said in a statement that the company had obtained the export control directive to suspend all foreign nationals’ access to Fable 5 and Mythos 5, without providing specific details regarding the national security concern.
The company said it believes Anthropic’s understanding is that the government has a way of bypassing or “jailbreaking” security measures that would prevent Fable 5 from being used to detect vulnerabilities in the software.
The order comes as a previous dispute between Trump administration officials and IPO-bound Anthropic showed signs of easing in some parts of the U.S. government.
Anthropic’s relationship with the government broke down this year after the US military refused to allow it to use AI models for domestic surveillance and fully autonomous weapons systems. The government responded by placing Anthropic on its supply chain blacklist, which comes into effect later in the year.
The action also marks a major escalation in U.S. efforts to stop foreign adversaries’ AI capabilities. For years, U.S. export controls have focused on the chips and tools that power AI, rather than restricting foreign access to AI.
Anthropic said the government had provided it with “only verbal evidence of a potential narrow-scope, non-universal jailbreak.”
“We disagree that a small finding of potential jailbreak should be a reason for a recall of a commercial model that has been applied to hundreds of millions of people,” the company said.
The government directive and Anthropic’s response underscore growing tensions between AI developers and regulators over how to assess risks from so-called “jailbreaks” or methods used to bypass model protections.
As recently as Wednesday, Anthropic had called for more U.S. oversight of AI, including the ability to block models that pose unacceptable risks. However, it was stated that the government’s action on Friday did not comply with the principles of fair and fact-based regulation.
Kirsten Davies, the Pentagon’s chief information officer, said in a post on X that the defense department supports prioritizing national security.
“Some things are more important than revenue cycles, clickbait and pre-IPO valuation. America first. Always,” Davies said.
Anthropic secretly filed for an initial public offering in the US last month, leapfrogging rival OpenAI in the race to reach public markets.
Earlier this week, Anthropic released an AI model called Claude Fable 5 that represents a new level of capability it calls the “Mythos class.” Anthropic said the model is accompanied by guardrails that prevent its use in risky areas such as cybersecurity, where some users complained it was “overly wide”.
Experts have said that if Mythos models fall into the wrong hands, they could significantly accelerate sophisticated cyberattacks, especially in industries such as banking that rely on complex, interconnected and often decades-old technology systems.
Anthropic said it was working with the US government, among others, on security ahead of the Fable launch, and that models from rival AI providers have shown a similar ability to uncover small bugs in code.
“The net effect of this order is that we are forced to 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.
Anthropic said it believed there was a “misunderstanding” and was working to restore access to the models as soon as possible.
A US official has confirmed that the commerce department has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by foreign nationals.
