World leaders including France’s Emmanuel Macron and India’s Narendra Modi expressed alarm at the G7 summit about the risk of the U.S. unilaterally cutting off access to American AI models, following the Trump administration’s recent export ban on Anthropic’s newest AI systems.
- G7 leaders caution U.S. AI cutoffs risk harming global economies and companies.
- Trump administration banned Anthropic model exports citing national security.
- Proposals emerge for trusted partner schemes to secure AI access internationally.
What happened
During the G7 Summit, French President Emmanuel Macron and Indian Prime Minister Narendra Modi voiced strong concerns about the U.S.'s ability to unilaterally disable access to American AI technologies. This alarm came soon after the Trump administration blocked exports of Anthropic's latest Mythos 5 and Fable 5 AI models on national security grounds. The order was influenced by reports from large technology firms warning that certain safety features in these models might be bypassed.
This decision effectively froze access to some of the most advanced AI services for international customers relying on Anthropic, raising fears that similar actions could disrupt other countries and businesses without warning or clear explanations. As a result, companies that depend on American AI infrastructure now face new uncertainties about the stability and continuity of AI-enabled products and services in their markets.
Why it matters
The incident underscores a major vulnerability in the global AI ecosystem: countries and companies outside the U.S. remain dependent on a handful of American tech providers for foundational AI technology. This dependence threatens the economic stability and national security of democracies, as access can be abruptly severed based on political or regulatory decisions in the U.S.
Experts and industry leaders highlight the importance of digital sovereignty—controlling the critical technologies that drive economic growth and security. The interruption of access to Anthropic’s models illustrates how that sovereignty is compromised when key technologies are centralized in one nation, emphasizing the need for diversified AI infrastructure and resilient international cooperation.
What to watch next
G7 leaders are exploring the development of a 'trusted partners' scheme aimed at granting select countries and companies reliable access to advanced AI models, bypassing unilateral U.S. restrictions. The initiative intends to foster a collaborative, open trade network for AI technology among democracies to strengthen defenses against geopolitical competitors such as China.
However, details remain unclear on how broadly the scheme would apply or how it might protect smaller AI startups worldwide. Policymakers and industry participants will also monitor Washington’s willingness to support this approach and expand access to crucial AI resources. The outcome will likely shape global AI governance frameworks and influence how nations balance innovation, security, and sovereignty moving forward.