OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei and Google DeepMind CEO Demis Hassabis are set to attend a lunch meeting during the G7 summit in Evian, France, on Wednesday. They will be joined by roughly a dozen other tech executives for talks centered on frontier AI risks, infrastructure and sovereignty.
CNBC reported that the session will take place alongside the summit. The Elysee Palace had previously said online child protection would also be part of the discussion. Beyond the three major AI labs, the guest list includes Mistral AI CEO Arthur Mensch, Cohere CEO Aidan Gomez, Meta's Alex Wang and Salesforce CEO Marc Benioff.
A seat at the table signals where influence sits
Jessica Brandt, senior fellow for technology and national security at the Council on Foreign Relations, told CNBC that heads of state now need cooperation from a small group of private-sector leaders actually building the technology if they want to make credible commitments on AI. In her view, the people invited into these negotiations reveal where power is concentrated.
That makes the presence of AI company leaders at a G7 meeting more than a symbolic industry appearance. It points to the growing geopolitical weight of a handful of firms as Europe also looks for ways to counterbalance US dominance in AI.
US export controls on Anthropic models remain unresolved
The summit comes as Anthropic is still in talks with the US government over export restrictions tied to its Fable 5 and Mythos 5 models. In June, US authorities cited national security concerns and ordered a suspension of access for all foreign users, saying the models could be jailbroken to identify software vulnerabilities and support cyberattacks.
The move marked a notable shift. It was the first time the United States applied export-control tools normally used for chips and weapons to a commercial AI model already used by hundreds of millions of people. Emerson Brooking, a senior fellow at the Atlantic Council, said the restrictions on Anthropic changed everything because Washington had shown it was willing to cut off even G7 countries and treaty allies from specific AI capabilities.
G7 leaders weigh a “trusted partners” framework
According to the report, G7 leaders are discussing a “trusted partners” framework that would selectively grant allies access to advanced models from US companies. Canadian Prime Minister Mark Carney said the restrictions highlight the danger of relying too heavily on a small number of American suppliers.
From model access to infrastructure control, the lunch meeting folds several issues into one conversation. The agenda now reaches well beyond technical safety and into the question of who gets access to advanced AI systems, and on what terms.