Demis Hassabis Calls for International AI Oversight as US Big Tech Slows and China Accelerates

The Quick Take

• Google DeepMind CEO Demis Hassabis publicly called for an international body to verify cutting-edge AI models before release, warning AGI could arrive within years with impacts 10 times the Industrial Revolution.
• Hassabis is meeting US policymakers to advocate for a pre-release AI certification structure modelled on financial self-regulatory organisations.
• Korean media coverage highlights the escalating AI hegemony competition between slowing US big tech and an accelerating Chinese AI sector, raising strategic technology sovereignty questions across Asia.

Google DeepMind CEO Demis Hassabis escalated his public advocacy for international AI governance, stating that artificial general intelligence capable of matching all human cognitive functions could arrive within years and would carry impacts ten times larger and faster than the Industrial Revolution. His proposal for a pre-release international verification body structured similarly to financial industry self-regulatory organisations represents a significant policy intervention from one of the most credible voices in advanced AI development. Korean coverage frames this in the context of intensifying US-China AI hegemony competition, where US big tech companies face competitive pressure from aggressive Chinese AI investment and rapid capability development.

The market implications for Asian technology investors are substantial. Korean semiconductor makers Samsung Electronics and SK Hynix supply high-bandwidth memory critical for both US and Chinese AI training infrastructure, positioning them as structural beneficiaries regardless of which country leads on model capability. However, any US export restriction expansion covering advanced memory chips would force Korean manufacturers to navigate difficult choices between US customer relationships and Chinese market access. Japanese technology companies and SoftBank portfolio companies face similar dual-exposure dynamics as AI investment flows from both US and Chinese hyperscalers into Asian supply chain partners.

The forward watch point is the outcome of Hassabis US policy meetings: if his pre-release verification body concept gains regulatory traction, it would benefit established AI labs including Google and Anthropic while imposing barriers on Chinese AI developers unlikely to submit to Western certification. The macro variable determining this AI competition outcome is US semiconductor export control policy, which remains the most powerful lever for maintaining the US-led AI capability advantage. Any tightening of controls on advanced chips to China will directly accelerate the bifurcation of the global AI supply chain that Korean and Japanese companies currently serve.

Synthesized from 2 sources.