U.S. authorities lifts restrictions on highly effective AI fashions, Anthropic says

Anthropic will quickly start restoring entry globally to its strongest AI fashions, Fable 5 and Mythos 5, after the U.S. government lifted a restriction on the place they may very well be launched, the corporate mentioned on Tuesday (June 30, 2026).

“We have obtained discover that the Division of Commerce has lifted export controls on Claude Fable 5 and Mythos 5,” Anthropic posted on X. “We’ll start restoring entry tomorrow.”

Earlier this month, U.S. authorities blocked entry to the fashions on nationwide safety grounds.

Simply 4 days in the past, the corporate mentioned it had obtained authorization from the federal government to permit a small group of American cybersecurity companies to entry Mythos 5.

Secretary of Commerce Howard Lutnick mentioned in a June 26 letter to the corporate that “Anthropic has labored with the U.S. authorities to handle dangers related to the Coated Fashions,” Politico reported.

Additionally Learn | Amazon voiced concerns about Anthropic AI models before U.S. government’s crackdown, source says​

The federal government abruptly pressured Anthropic to chop off entry to its two cutting-edge synthetic intelligence fashions on June 12 after discovering vulnerabilities within the safeguards put in place to stop misuse of the software.

Like Anthropic, rival AI lab OpenAI has additionally complied with Washington’s requests to limit its personal launch of a brand new, highly effective mannequin known as GPT-5.6 to a restricted set of authorized companions.

“This is not fairly the method that we predict is perfect,” OpenAI CEO Sam Altman mentioned Friday in a put up on X, alongside a proof of the GPT-5.6 launch.

Anthropic didn’t instantly reply to a request for touch upon Tuesday (June 30, 2026).

Earlier within the day, CIA Director John Ratcliffe in contrast the capabilities of probably the most superior synthetic intelligence fashions to nuclear weapons, in a tacit protection of the Trump administration’s latest onerous line on controlling the discharge of probably the most highly effective AI know-how.

“In conversations with most of the president’s different nationwide safety and financial safety advisors, we’re speaking concerning the influence of those frontier AI fashions,” Ratcliffe mentioned throughout a speech on the AWS summit in Washington.

“It might be… not misplaced to seek advice from their capabilities as akin to digital nuclear weapons,” Mr. Ratcliffe mentioned.