World leaders want American AI. They just don't want America to be able to turn it off

French President Emmanuel Macron and Indian Prime Minister Narendra Modi raised concerns at the G7 Summit on Wednesday that the United States could suddenly restrict access to leading American artificial intelligence models. Macron told G7 leaders and AI company executives that such a cutoff could damage the economies of European users as well as the AI companies involved. The remarks followed the Trump administration's decision to block Anthropic from exporting its Mythos 5 and Fable 5 models due to national security concerns after Amazon alerted the White House that safety features might be circumvented. Cybersecurity experts have noted that similar capabilities exist in models from other providers that remain available. Prime Minister Modi expressed worry over the Anthropic restriction and stated that democratic nations require unrestricted access to advanced AI models to safeguard critical infrastructure. Aidan Gomez, co-founder and chief executive of Canadian firm Cohere, said in a statement that the restriction highlights the risks of depending on a few large technology companies and emphasized the importance of digital sovereignty in controlling foundational technology. G7 leaders also considered a trusted partners arrangement that would allow non-U.S. countries access to advanced models from companies including Anthropic and OpenAI. The scheme aims to create an open trade network that avoids U.S. export limits while enabling development of stronger defenses against competitors such as China. It remains unclear how broadly the arrangement would apply. Macron suggested that Washington should support the scheme and provide wider access to the Mythos model. He observed that buyers would hesitate to purchase U.S. AI access if it could be revoked without notice.