Top AI Bosses Issue Warning to G7 Leaders Over Control Timelines

The chiefs of the world's prominent technology firms issued a direct plea to world leaders during a working lunch at the conclusion of the Group of Seven (G7) summit in the French Alps. They urged governments to establish frameworks to govern Artificial Intelligence (AI).

The high-level discussions took place in Evian-les-Bains, where the heads of three dominant American firms joined state leaders. OpenAI Chief Executive Officer (CEO) Sam Altman, Google DeepMind head Demis Hassabis, and Anthropic leader Dario Amodei all participated in the session.

Their presence came at a friction-filled moment for global digital infrastructure. Just days earlier, the administration of United States (US) President Donald Trump compelled Anthropic to restrict access to its most advanced models, raising widespread diplomatic anxieties over Washington controlling a unilateral technological kill switch.

According to individuals familiar with the closed-door talks, attendees avoided direct mention of the enforcement action. Instead, the tech executives focused their addresses on international regulation and the immediate timeline for state intervention.

The three corporate leaders delivered stark warnings regarding the rapid advancement of their systems. They agreed that governments have a narrow window to establish control, though their specific timelines on capabilities diverged.

Hassabis outlined a development window of three to five years. Amodei estimated that the sector is merely one or two years away from creating models that surpass human capabilities across all fields, which would completely upend global digital systems.

Altman supported this compressed outlook, stating he expects systems of astonishing power within the next two years. The executives omitted discussions on domestic employment or economic inequality, which are typical topics in public policy forums.

Instead, they focused on national security risks, highlighting threats related to cyber warfare, bioterrorism, and nuclear proliferation. Amodei warned that, if current trajectories continue, AI will become the foundational basis for both military and economic dominance among nation-states.

This warning carries geopolitical weight for international delegates, because the leading tech laboratories remain consolidated within the US. To address these threats, the executives called for an international governance architecture, though they disagreed on who should hold the keys.

Hassabis advocated for a global technical standards body supported by the leading corporate laboratories. He suggested this entity could adjust safety baselines every quarter, as new risks appear in the underlying technology.

Amodei proposed a narrower alliance, calling for a coalition led by Washington alongside democratic nations to control access to advanced models and isolate adversaries like China.

Altman rejected the restriction of access, arguing that once baseline guardrails are established, global policy should prioritize human liberty. He argued against concentrating control in a small number of entities, describing that centralization as an insidious threat.

The summit also featured discussions among spouses of the global leaders. French First Lady Brigitte Macron led a delegation, including Kenya First Lady Rachel Kimetto, to examine the specific risks that unregulated automated systems pose to children.

While the primary G7 sessions focused heavily on traditional geopolitical conflicts like the war in Ukraine, the sudden insertion of the technology executives into the final ministerial lunch highlights the rising strategic importance of digital infrastructure.

The sharp policy divisions among the corporate chiefs left world leaders without a unified framework, but it remains uncertain how international bodies will reconcile the competing demands for strict state control.

The event underscored that the future of global tech infrastructure is no longer just a commercial concern, but a matter of national defense.