US restrictions on Anthropic models could trigger global AI arms race

The US Commerce Department issued an export-control directive on June 12 that immediately suspended foreign access to Anthropic's Fable 5 and Mythos 5 AI models, escalating a standoff that began when Anthropic refused Pentagon demands for unrestricted use of its technology. The move sets a precedent for using export controls on AI model weights, potentially triggering a global AI arms race.

US restrictions on Anthropic models could trigger global AI arms race The Commerce Department's export-control order on Anthropic's newest AI models marks a dramatic escalation in the standoff between Washington and one of its most prominent AI companies. The US government just cut off the rest of the world from Anthropic’s most powerful AI systems. On June 12, the Commerce Department issued an export-control directive that immediately suspended foreign access to Anthropic’s Fable 5 and Mythos 5 models, which had launched just three days earlier on June 9. How we got here The roots trace back to February 27, when a presidential directive led to the Pentagon designating Anthropic as a “supply-chain risk.” The reason: Anthropic refused the Pentagon’s demands for unrestricted use of its technology, particularly for applications like autonomous weapons systems and domestic surveillance. Following that designation, federal agencies were directed to stop using Anthropic’s models entirely. The export-control order in June represents the next logical escalation, a tool typically reserved for semiconductors and defense technology now being wielded against AI model weights. Anthropic CEO Dario Amodei has actually been one of the loudest voices in favor of strong export controls on advanced AI. He’s publicly advocated for careful regulation of advanced chips and model weights. The irony is that the government is now using those same export-control mechanisms against his own company, not because of what Anthropic did with its technology, but because of what it refused to do. The models at the center of it all Fable 5 and Mythos 5 represent Anthropic’s latest generation of AI systems, and they’re architecturally distinct in one critical way. Fable 5 includes safety classifiers, the kind of built-in guardrails designed to prevent misuse. Mythos 5 does not. The ban applies to all foreign nationals, including Anthropic’s own employees who hold non-US citizenship. Tech companies routinely employ engineers and researchers from around the world, and freezing them out of access to the company’s own products creates immediate operational headaches alongside the geopolitical ones. The US has used similar export restrictions on advanced semiconductor chips, most notably against China. But applying this framework to AI model weights is a newer and more aggressive posture. Chips are physical objects that cross borders in shipping containers. Models are software that can be accessed through an API from anywhere on earth. What this means for the global AI landscape The precedent being set here extends well beyond Anthropic. Any American AI company that pushes back on government demands for unrestricted access now knows what the playbook looks like: supply-chain risk designation, federal contract exclusion, and export controls. Crypto-native media have mainly focused on the government-AI firm clash rather than impacts on token markets, reflecting a legal-centric narrative rather than a financial one. No specific tokens or protocols have been directly affected by the Anthropic restrictions. Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy https://cryptobriefing.com/editorial-policy/ .