Top AI CEOs meet with G7 leaders amid European sovereignty fears

AI Chiefs Altman, Amodei, and Hassabis Join World Leaders at G7 Summit in France

The heads of the three most powerful frontier AI labs — OpenAI’s Sam Altman, Anthropic’s Dario Amodei, and Google DeepMind’s Demis Hassabis — gathered Wednesday with world leaders at the G7 summit in Evian-les-Bains, France, marking the first time all three rival CEOs have appeared together before heads of state at the annual gathering.sfgate

The executives joined a working lunch themed around “ensuring a safe, rapid and effective deployment of artificial intelligence” on the final day of the three-day summit, which had been dominated by discussions on the wars in Ukraine and Iran.apnews

European Sovereignty Fears Take Center Stage

The AI session unfolded against a backdrop of intensifying European anxiety over American dominance of the technology. Days before the summit, Washington suspended foreign access to Anthropic’s most advanced AI models, citing national security concerns — an episode that confirmed what European officials have described as a potential American “kill switch” over critical technology infrastructure.euronews

“Tech is more and more becoming a strategic asset. Europe must be able to act on its own terms,” said Marie-Anne Regnier, a French government official, according to Euronews. The European Commission had already unveiled a tech sovereignty package this month aimed at boosting homegrown AI capabilities.wsls

Reuters reported that G7 leaders on Tuesday discussed a potential “trusted partners” scheme that would grant allied nations exemptions from the U.S. restrictions on frontier AI models. The EU is also seeking access to Anthropic’s Mythos model, designed to detect flaws in computer code, in order to study its implications for European cybersecurity.thestar

Child Safety and a Broader Tech Agenda

Beyond the headline AI session, Mistral AI’s Arthur Mensch attended a separate discussion focused on protecting children in the digital sphere. France’s President Emmanuel Macron had made online child safety a centerpiece of France’s G7 presidency, declaring in an Instagram video that the summit would focus on how to “improve cyber security and protect our children and our democracies”.rte

G7 digital ministers had laid the groundwork in late May in Paris, agreeing on shared principles for age verification, tackling illegal content, and addressing risks to children from AI chatbots.ukpol

Also attending the AI discussions were the heads of smaller labs representing each G7 nation, including Canada’s Cohere, Germany’s Black Forest Labs, Italy’s Domyn, Japan’s Sakana AI, and the U.K.’s Synthesia. Cohere CEO Aidan Gomez said his company’s goal at the summit was “to expand our sovereign AI ecosystem partnerships beyond Canada and Germany to include all G7 nations”.sfgate