Anthropic Nearing Deal to Remove Export Controls on AI Models

Anthropic is approaching an agreement with the Trump administration to lift federal export restrictions on two of its advanced AI systems, according to sources familiar with the ongoing negotiations.

Commerce Secretary Howard Lutnick has been working to address security concerns that would enable the removal of export controls currently imposed on Anthropic's Fable 5 and Mythos 5 models. The restrictions could be removed once officials throughout the administration approve the arrangement, the sources indicated.

The development follows weeks of discussions between Anthropic and government officials focused on the security protocols surrounding these AI systems. Bloomberg first reported the details of these negotiations.

Why it matters

Export controls on AI models represent a significant regulatory tool the U.S. government uses to manage the international distribution of advanced artificial intelligence technology. When such restrictions are placed on commercial AI systems, they can limit a company's ability to serve international customers and compete in global markets. The potential lifting of these controls suggests Anthropic has satisfied government concerns about how its systems might be used or accessed abroad—a precedent that could influence how other AI companies approach security commitments with federal regulators.

Security negotiations at the center

The talks between Anthropic and the Commerce Department have centered on establishing adequate safeguards for the two AI models. While the specific security measures under discussion have not been disclosed, export controls on AI systems typically relate to concerns about technology transfer, potential misuse by foreign actors, or national security implications.

The involvement of multiple administration officials in the approval process indicates the decision carries weight beyond the Commerce Department alone. Cross-agency coordination on AI export policy has become increasingly common as the federal government works to balance innovation with security considerations.

Implications for AI regulation

The case highlights the evolving relationship between AI companies and federal regulators as advanced models become more capable. Anthropic, backed by significant investment from Amazon and Google, has positioned itself as a safety-focused AI developer. The company's willingness to engage in extended negotiations over export controls demonstrates the regulatory complexity now facing major AI labs.

The outcome of these discussions could establish a framework for how similar situations are handled with other AI companies facing export restrictions. As AI capabilities continue to advance, the question of which systems require export controls—and under what conditions those controls can be lifted—will likely remain a key policy challenge.

These details were first reported by Bloomberg.