The models had been on the list for months. The Trump administration has lifted export restrictions on Anthropic’s flagship Claude models, Fable 5 and Mythos 5, according to CNBC, clearing the way for the AI company to sell its most advanced technology across international markets.
The controls were imposed earlier this year, when officials placed the two models on a restricted list over what they described as national security risk. The designation required Anthropic to obtain licenses before selling the models outside the United States, a process that chilled overseas deals and gave customers in Europe, Asia and the Middle East reasons to look elsewhere.
The removal is the first major regulatory win for Anthropic as it prepares for an initial public offering, a process that people familiar with the matter said is in its final stages. The company has been holding talks with banks and advisers for months, and executives had privately complained that the export restrictions would hang over the listing like a discount sticker: investors value a model on how many markets it can reach, and Fable 5 and Mythos 5 were effectively confined to one.
What the decision unlocks is immediate. Anthropic can now sell its flagship models to enterprises and governments abroad without case-by-case approval. That matters most in Europe, where large companies have been cautious about adopting American AI products under tight data rules, and in the Gulf states, where sovereign funds have been writing large checks for AI infrastructure. A person close to the company said the sales pipeline has already started moving, with several multi-year enterprise contracts in advanced negotiation.
The financial stakes are substantial. Anthropic closed a $65 billion funding round in May that drew Samsung as a strategic investor, according to reports, and the company’s valuation has been climbing alongside its revenue. Analysts said the export clearance removes one of the biggest overhangs on that valuation, because restricted models are worth measurably less than unrestricted ones. “The market prices AI companies on addressable market,” one analyst said. “Anthropic just got its addressable market back.”
The decision also sharpens the contrast with OpenAI, whose latest model, GPT-5.6, remains in government review after the White House asked for a delay. Two of the most valuable private AI companies now face different regulatory weather: Anthropic cleared, OpenAI negotiating. People familiar with both companies said the divergence is not lost on investors, who are watching whether clearance becomes a pattern or a one-off.
The timing is politically delicate. Some lawmakers have pushed for tighter controls on frontier AI, arguing that the technology is too powerful to sell freely, and the administration’s decision to relax restrictions on two leading models is likely to draw scrutiny on Capitol Hill. Anthropic has positioned itself as the safety-conscious alternative in the AI industry, a reputation that may have helped its case. The company has built its public identity around cautious deployment, and officials weighing the export decision were said to have cited the company’s safety record in their deliberations.
The operational effects of the restriction were felt across the company. Every export license application meant documentation, legal review and waiting periods that could stretch for months, and in several cases, according to people familiar with the sales process, foreign customers simply walked away rather than wait. Competitors without U.S. export exposure, including European and Chinese labs, were free to sell comparable models without the paperwork, and Anthropic’s sales teams watched deals drift to them.
The clearance also matters for the company’s positioning in the safety debate. Anthropic has built its public identity around cautious deployment, and officials weighing the export decision were said to have cited the company’s safety record in their deliberations. Executives hope the decision establishes a precedent that makes future reviews faster, since the government now has a track record of clearing Anthropic’s models for sale.
The competitive effects reach beyond the two companies. Open-weight models from other developers have been circulating internationally without restriction, and U.S. officials have worried that export limits on American frontier models simply pushed foreign buyers toward Chinese alternatives. Lifting the restrictions on Anthropic’s models gives American AI a firmer position in the global market at a moment when the competition with China’s leading labs is intensifying.
Not everything is settled. The clearance applies to the current generation of models, and future releases will be reviewed on their own terms. Anthropic has new models in training, and executives are expected to press for a faster, more predictable review process as the company grows. The government, for its part, has kept the machinery of oversight in place, and officials have said the decision reflects the specific characteristics of these models rather than a general relaxation of policy.
For Anthropic, the immediate effect is simpler: the company can now sell its best products everywhere. With an IPO approaching, with a $65 billion funding round behind it and with export controls lifted, the company enters the public markets with its main regulatory problem resolved. The models are off the list; the question now is how many customers line up.
