The U.S. Department of Commerce has lifted export controls on Anthropic’s Claude Fable 5 and Mythos 5 AI models, bringing an end to a dramatic standoff with the Trump administration over national security concerns. This decision allows Anthropic to re-enable access to its advanced AI for global users and select partners, alleviating industry fears about hindering American AI innovation in the face of growing Chinese competition. The resolution follows intense negotiations and comes amidst broader uncertainty regarding AI regulation from the U.S. government.
In a significant development for the artificial intelligence industry, Anthropic announced on Tuesday that the U.S. Department of Commerce has officially lifted export controls on its highly anticipated AI models, Claude Fable 5 and Mythos 5. This decision brings an end to a tense and dramatic standoff between the prominent AI company and the Trump administration.
“We're grateful to our users for their patience, and to everyone who worked with us on redeploying the models,” Anthropic stated in a post on X, acknowledging the disruption caused by the temporary ban. The company had previously disabled access to Fable 5 and Mythos 5 in mid-June to comply with a government directive citing “national security authorities.” This directive had instructed Anthropic to suspend all access by “any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.”
With the controls lifted, Fable 5 will once again be available to global users on the Claude platform, Claude.AI, and Claude Code starting Wednesday. Anthropic also confirmed that the AI model will be included for up to 50% of weekly usage limits through July 7 for users of Pro, Max, Team, and selected enterprise plans.
Furthermore, access to Mythos 5 has been restored for some U.S. organizations, following government approval granted on June 26. Anthropic plans to continue collaborating with the U.S. government to expand access to Mythos 5 for more domestic and international partners through its Glasswing program, a cybersecurity initiative offering advanced AI models for defensive security testing. Additionally, Fable 5 will be re-enabled on major cloud platforms including Amazon Web Services, Google Cloud, and Microsoft Foundry as soon as possible.
The government’s earlier crackdown on Anthropic coincided with a rapid emergence of capable and significantly more affordable Chinese open-source AI models. This situation prompted concerns among tech executives and investors that the U.S. administration’s restrictions were inadvertently providing valuable time for Chinese developers to close the AI innovation gap.
This resolution comes just days after U.S. Commerce Secretary Howard Lutnick gave Anthropic initial permission to release Mythos 5 to a select group of companies and federal agencies. In a letter viewed by CNBC, Lutnick stated that “appropriate safeguards” were in place for “trusted partners” to access the model. On Tuesday, Lutnick posted on X, reaffirming, “Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America's leadership in AI.”
Days prior to the imposition of export controls, Anthropic had launched these models, touting their state-of-the-art capabilities across various industry benchmarks. Fable 5, in particular, represented Anthropic's first public release of such an advanced offering.
Anthropic, which has faced a contentious relationship with the White House this year, swiftly dispatched a team to Washington, D.C., to negotiate with the Trump administration after receiving the export control directive on June 12. Tom Brown, an Anthropic co-founder, reportedly led these critical negotiations, stepping into a role previously handled by CEO Dario Amodei, who had been a target of the administration due to his outspoken views on AI safety and his support for Kamala Harris in the 2024 presidential election. Notably, Lutnick's letter on Friday was addressed to Brown, not Amodei.
The incident has sparked considerable confusion across the AI industry, with leaders expressing uncertainty about the government’s evolving regulatory approach and who precisely is influencing presidential decisions, especially after David Sacks vacated the role of crypto and AI czar earlier in the year.
The administration has signaled a more active role in AI regulation, with President Donald Trump signing an AI executive order in early June. This order, light on specific details, requested AI developers voluntarily submit their models to the government for capability assessment prior to full release, giving federal agencies 60 days to establish the necessary frameworks. This has proven challenging for other AI companies, including OpenAI, attempting model launches in the interim.
For instance, OpenAI recently announced three new models, including GPT-5.6, and confirmed its compliance with the government’s request to initially limit their rollout to a “small group of trusted partners.” OpenAI previewed the models' capabilities and shared its launch plans with the government. However, the company voiced a sentiment shared by many in the industry: “We don't believe this kind of government access process should become the long-term default. It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them.”
— CNBC's Justina Lee contributed to this story.
