The United States has barred G7 countries from accessing Anthropic’s most advanced artificial intelligence models, citing national security concerns. The restriction, imposed last week under export controls announced on June 12, prevents use of Anthropic’s Mythos and Fable AI models by foreign nationals and entities outside the United States.

UK Prime Minister Keir Starmer reportedly sought an exemption allowing British citizens and companies to access the technologies, requesting a “carve-out” to the U.S. ban. However, a Trump administration official described granting any exceptions, even to allied G7 members, as “completely illogical,” emphasizing the priority of safeguarding the technology.

According to sources, the U.S. government is collaborating with Anthropic to ensure the safety and security of the AI models globally. Officials conveyed that the actions reflect America’s intent to protect sensitive technology rather than restrict its development outright.

The White House is engaged in discussions with Anthropic to find a resolution that addresses the concerns behind the export controls. A senior administration official indicated that negotiations might proceed directly with Anthropic’s CEO, Dario Amodei. Anthropic has not commented publicly on the situation.

The export controls announced on June 12 represent a broader U.S. effort to regulate the dissemination of advanced AI technologies to foreign entities amid heightened scrutiny over potential risks related to national security. The restrictions currently bar any use of Anthropic’s newest AI infrastructure by individuals or companies based outside the U.S. or by non-U.S. citizens, effectively cutting off access to G7 partners, despite the bloc’s collaborative ties.

Efforts by the UK government to secure an exemption highlight tensions between allied countries over technology control policies. The U.S. stance underscores its commitment to prevent sensitive AI capabilities from falling into hands it deems risky, even within established alliances. The outcome of ongoing talks with Anthropic may influence how allied nations access and deploy cutting-edge AI tools in the future.