Dario Amodei, co-founder and CEO of Anthropic, attends the AI Impact Summit in New Delhi, India on Thursday, February 19, 2026.
Prakash Singh | Bloomberg | Getty Images
Anthropic announced Tuesday that the U.S. Department of Commerce has lifted export restrictions on its Claude Fabre 5 and Mythos 5 models, ending the latest dramatic standoff between the artificial intelligence company and the Trump administration.
“We would like to thank our users for their patience and everyone who helped us redeploy the model,” Anthropic said in a post on X.
Anthropic disabled access to Fable 5 and Mythos 5 in mid-June to comply with export control directives from the government citing “national security authorities.” The company said it was directed to terminate all access “by foreign nationals, including Anthropic’s foreign employees, whether in the United States or abroad.”
The government’s crackdown on Anthropic coincided with the rapid rise of Chinese open-source models that have proven to be nearly as capable and significantly cheaper than some of America’s most powerful models. With the Trump administration restricting the rollout of Anthropic’s latest models, many technology executives and investors have expressed concern that Chinese developers are being given precious time to catch up.
Tuesday’s announcement comes days after U.S. Secretary of Commerce Howard Lutnick granted Anthropic permission to release Mythos 5 to a select group of companies and federal agencies. In a letter to the company seen by CNBC, Lutnick said he determined that “appropriate safeguards” were in place to allow access to the models to certain “trusted partners.”
“Over the past two weeks, we have worked closely with Anthropic to ensure alignment across the U.S. government and analyze and approve Fable 5 to strengthen U.S. leadership in AI,” Lutnick wrote in a post on X on Tuesday.
Anthropic announced Wednesday that it will begin restoring access to Fable 5.
Days before export restrictions, Anthropic launched these models and touted them as cutting edge across various industry benchmarks. Fable 5 in particular marks the first time Anthropic has released such an advanced product to the public.
The company has been at odds with the White House for much of this year, but after receiving the export control directive on June 12, it rushed to Washington, D.C., to negotiate with the Trump administration. Anthropic remained tight-lipped over the next few weeks, sharing minimal public updates on the status of negotiations.
Anthropic co-founder Tom Brown is reportedly leading negotiations with the Trump administration on behalf of Anthropic CEO Dario Amodei. Amodei has been targeted by the administration for his outspoken views on the safety of AI and his vocal support for Kamala Harris in the 2024 presidential election.
Mr. Lutnick’s letter on Friday was addressed to Mr. Brown, not Mr. Amodei.
The incident has caused confusion across the AI industry, especially since David Sachs abdicated his role as crypto and AI czar earlier this year, with leaders saying they are in the dark about the government’s regulatory pursuits and don’t know who has decision-making power or influence over the president.
The administration has indicated it intends to take a more active role in AI regulation, with President Donald Trump signing an AI executive order in early June. The order was light on specific details, asking AI developers to voluntarily submit their models to the government to evaluate their functionality before full release. This gave federal agencies a 60-day grace period to establish relevant frameworks and processes.
This has been difficult for AI companies, including OpenAI, who have been trying to launch models in the meantime.
OpenAI announced three new models on Friday, saying it was responding to government requests to initially limit deployment to a “small group of trusted partners.” OpenAI said it previewed the model’s capabilities, including GPT-5.6, and shared its plans with the government ahead of Friday’s launch.
“We do not believe this type of government access process should become the long-term default,” OpenAI said in a post Friday. “We secure the best tools from the users, developers, enterprises, cyber defenders, and global partners who need them.”
WATCH: Why OpenAI and Anthropic are rushing IPOs amid concerns about AI premium decline
