AI Executives Join G7 as US and China Diverge on Tech Governance

Private AI companies now shape diplomatic agendas once reserved for governments alone. At the G7 summit in Évian, France, executives from OpenAI, Anthropic, Google DeepMind and a dozen other firms joined national leaders for closed-door talks on frontier risks, infrastructure and child safety online.

The presence of these CEOs marks a concrete shift. Policy commitments on advanced AI now require input from the small group of labs that control the most capable systems. France’s Élysée Palace confirmed the lunch agenda would cover sovereign AI capabilities and cross-border infrastructure needs.

Separate from the summit, Chinese officials released a white paper outlining plans for a global AI cooperation body open to all countries. Foreign Minister Wang Yi stated that technology must serve human needs and criticized exclusive arrangements. Vice chair Zhao Haibing of China’s top economic agency called for cooperation through BRICS and the Shanghai Cooperation Organization instead.

The United States has signaled tighter controls on advanced models for non-allied nations. A Reuters report described discussions among G7 members about granting “trusted partners” preferential access; CNBC stated it could not independently confirm those details. Chinese representatives contrasted their emphasis on downloadable, low-cost models with subscription-based Western offerings.

No binding agreements emerged from the G7 sessions. Participants instead discussed possible voluntary pledges on cyber and biological risks. The gap between Washington’s security-focused restrictions and Beijing’s inclusive framework remains unresolved, leaving smaller nations to navigate competing standards.