Anthropic announced on Tuesday that it will begin restoring global access to its most powerful artificial intelligence models, Fable 5 and Mythos 5, after the US government lifted restrictions that had limited their release. The company confirmed via a post on X that the Department of Commerce had removed export controls on the two models, with access restoration starting the following day.
Background of Restrictions
The Trump administration had invoked national security concerns over the past two weeks to limit major US tech companies from releasing advanced AI models. For Anthropic, this meant cutting off access to Fable 5 and Mythos 5 on June 12 after vulnerabilities were discovered in the safeguards designed to prevent misuse. Some researchers feared these models could be exploited to bypass cybersecurity measures.
Just four days before the full lift, Anthropic had received authorization to allow a small group of American cybersecurity firms to access Mythos 5. Secretary of Commerce Howard Lutnick, in a June 26 letter to the company, stated that “Anthropic has worked with the US government to address risks associated with the Covered Models,” as reported by Politico.
Government Decision
On Tuesday, Lutnick informed Anthropic in a letter that the Trump administration had “withdrawn” its previous restrictions on the release of the company’s models. The letter indicated satisfaction that Anthropic had “taken steps in close coordination with the US government to address the risks associated with Claude Mythos 5 and Claude Fable 5.”
This move follows a broader pattern of government oversight on advanced AI. Rival lab OpenAI also complied with Washington’s requests to restrict its new model, GPT-5.6, to a limited set of approved partners. OpenAI CEO Sam Altman commented on Friday via X, saying, “This isn’t quite the process that we think is optimal,” alongside an explanation of the GPT-5.6 launch.
Executive Order and Cooperation
The Trump administration issued an executive order on June 2, calling for federal actions on AI and cybersecurity over the subsequent two months. This included creating a voluntary “framework” for private companies like Anthropic and OpenAI to test and release powerful “frontier” AI models in collaboration with the government.
Susie Wiles, the president’s chief of staff, posted on X Tuesday expressing gratitude for cooperation from tech companies, though she did not name any. She said, “My gratitude to companies across industries who continue to work closely with the White House to implement the President’s executive order on AI and cybersecurity. This includes excellent work around advanced model access and guardrail testing and security.”
Comparison to Nuclear Weapons
Earlier in the day, CIA Director John Ratcliffe compared the capabilities of advanced AI models to nuclear weapons during a speech at the AWS summit in Washington. He defended the administration’s hard line on controlling AI release, stating, “In conversations with many of the president’s other national security and economic security advisers, we’re talking about the impact of these frontier AI models. It would be...not misplaced to refer to their capabilities as akin to digital nuclear weapons.”
Anthropic did not immediately respond to a request for comment on Tuesday.
