Anthropic said Tuesday it will soon begin restoring global access to its most powerful AI models, Fable 5 and Mythos 5, after the U.S. government lifted restrictions on their distribution, UNN reports citing AFP.
Details
Over the past several weeks, the Trump administration has cited national security concerns to restrict the ability of major U.S. technology companies to release advanced models, including Anthropic's models, which some researchers believed could be used to bypass cybersecurity measures.
"We have received notification that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5. We will begin restoring access tomorrow," Anthropic wrote on X.
Just four days ago, the company said it had received permission from the government to provide access to Mythos 5 to a small group of U.S. cybersecurity firms.
As Politico reported, U.S. Commerce Secretary Howard Lutnick said in a June 26 letter to the company that "Anthropic has worked with the U.S. government to address risks associated with the covered models."
On June 12, the U.S. government abruptly forced Anthropic to suspend access to two advanced AI models after discovering vulnerabilities in safeguards designed to prevent misuse of the tool.
On Tuesday, Lutnick informed Anthropic in a letter that the Trump administration had "rescinded" previously imposed restrictions on the release of the company's models, Politico reported.
The letter indicated that the Trump administration is satisfied, at least for now, that Anthropic "has taken steps in close coordination with the U.S. government to address risks associated with Claude Mythos 5 and Claude Fable 5."
Like Anthropic, rival AI lab OpenAI also complied with Washington's request to limit the release of its new powerful model GPT-5.6 to a limited circle of approved partners.
"This is not exactly the process we consider optimal," OpenAI CEO Sam Altman said Friday in a post on X, accompanying it with an explanation of the GPT-5.6 launch.
In a blog post published Tuesday evening, Anthropic called for the development of standardized frameworks for assessing and responding to critical vulnerabilities in advanced models.
The San Francisco-based AI lab will collaborate with Amazon, Microsoft, Google and others in this direction.
"This problem will intensify in the coming months as more models with powerful cybersecurity (and other) capabilities are trained, evaluated, and released," the blog post said.
Austria urges Europe to host Anthropic after US restricts access to AI29.06.26, 08:48
On June 2, the Trump administration issued an executive order calling on the federal government to take a series of steps in AI and cybersecurity over the next two months, including creating voluntary "frameworks" for private companies such as Anthropic and OpenAI to test and release their powerful "frontier" AI models in cooperation with the government.
Susie Wiles, White House Chief of Staff, posted on X Tuesday that the Trump administration is grateful for cooperation from technology companies, though she did not name any of them.
"I am grateful to companies across various industries that continue to work closely with the White House to implement the president's executive order on AI and cybersecurity," Wiles said. "This includes excellent work on ensuring access to advanced models, testing, and security."
Earlier that day, CIA Director John Ratcliffe compared the capabilities of advanced AI models to nuclear weapons, thereby indirectly defending the Trump administration's recent tough stance on controlling the spread of the most powerful AI technologies.
"In conversations with many other presidential advisors on national security and economic security, we are discussing the impact of these advanced AI models," Ratcliffe said during a speech at the AWS Summit in Washington.
"It would be... entirely appropriate to compare their capabilities to digital nuclear weapons," Ratcliffe added.