French President Emmanuel Macron urged the world's wealthy democracies on Wednesday to create a shared framework for regulating advanced artificial intelligence, as tensions over U.S. restrictions on AI model access injected urgency into a G7 summit in Evian-les-Bains, France. The push came hours after OpenAI CEO Sam Altman told the same gathering that AI safety standards should not be left to tech companies alone.
Anthropic directive fuels European anxiety
The White House's move last week to block foreign nationals from using Anthropic's newest AI models - Fable 5 and Mythos 5 - underscored how quickly a nation can be severed from critical technology. Macron called the U.S. directive a "strictly nationalist" reaction, even as he acknowledged it showed American officials understood the dangers of frontier AI systems. He warned that American firms could lose value if they cut off access abruptly.
The episode forced Anthropic to take the models offline to comply. The company said it did not believe the government's security concerns justified the steps. For European officials, the incident crystallized long-standing fears about dependence on U.S. AI giants.
Calls for international AI governance
Altman told the working lunch, attended by G7 leaders and more than a dozen AI executives, that an "international forum" is needed to establish globally accepted testing standards and provide expert risk analysis. "We need an international forum for discussion that establishes globally accepted standards for testing, provides expert and impartial analysis of capabilities and risks, and serves as a venue for cooperation among nations," he said. He stressed that the technology's future must be shaped by people and democratic institutions, "not just by the companies building the most capable systems."
Macron backed the idea of closer government-to-government coordination on security and cybersecurity. For policy professionals, these proposals signal a shift toward structured multilateral oversight, similar to frameworks explored in AI for Policy Makers. He also pleaded with the U.S. not to keep advanced AI to itself, arguing that democratic countries ultimately want to prevent authoritarian regimes from gaining access.
Europe pushes sovereignty as insurance
Before the Anthropic dispute, the European Commission had already rolled out a tech sovereignty package to boost homegrown AI. Macron said France will increase funding for its own AI industry so it is not left behind if international cooperation collapses. Zach Meyers, research director at the Brussels-based think tank CERRE, said the U.S. action showed how Europe and other allies "can be put in an extremely vulnerable position" when cut off from advanced models.
Aidan Gomez, CEO of Canada's Cohere AI, told the AP that the consensus at the G7 was "we need something" on joint AI governance. He said democracies should ensure the G7 does not merely produce the most capable AI but also "the second most capable AI," acknowledging the current duopoly of the U.S. and China. These developments align with growing public-sector interest in AI for Government resilience.
Why this matters for government professionals
The G7 discussions and the Anthropic restriction expose a hard reality: nations without domestic AI capacity or binding multilateral agreements can be cut off from strategic tools overnight. For government officials, the immediate takeaway is that regulatory cooperation and investment in local AI infrastructure are not abstract policy goals - they are operational necessities. The push for an international forum, the European tech sovereignty drive, and Macron's plea for partnership all point to a future where diplomatic and technical planning must move in lockstep. Professionals shaping procurement, national security, and digital strategy will need to track these fast-moving cross-border negotiations and their legal and practical consequences.
Your membership also unlocks:
