At the G7 Summit, world leaders expressed concern that the U.S. could abruptly cut off access to its advanced AI models, potentially harming economies and undermining tech firms’ stability.

The risk was highlighted by the recent decision of the Trump administration to block Anthropic’s export on national security grounds, despite arguments from cybersecurity experts about the availability of similar capabilities in other models. This episode underscores the vulnerability for international companies and governments relying on U.S. AI infrastructure.

Prime Minister Modi echoed fears that such access can be revoked overnight without explanation, emphasizing the need for democratic nations to maintain unfettered access to top AI technologies to protect critical infrastructure. Experts argue that digital sovereignty is crucial for long-term economic security and national interests, as control over foundational technology will shape global powers.

In response, G7 leaders discussed a “trusted partners” scheme aimed at granting non-U.S. nations access to advanced AI models from firms like Anthropic and OpenAI, bypassing current U.S. restrictions. However, the scope of this scheme remains unclear, leaving many uncertain about how it will address sudden disruptions in AI access.

As Europe and other countries strive for AI sovereignty, they face challenges when faced with the superior performance of American models and the fear of being left behind if their access is jeopardized by U.S. decisions.