Breaking

Demis Hassabis Backs Call to Pace Frontier AI Development

September 13, 2026 at 11:15 EDT

Foundation Models
AI Agents
Policy & Regulation

Demis Hassabis, Google DeepMind co-founder and Alphabet chief scientist, has endorsed a call to slow the pace of frontier AI development, aligning one of the industry's leading builders with Anthropic CEO Dario Amodei's proposal to give safety work more time to catch up. Hassabis also indicated that the proposal's details still need to be worked through, while backing its overall direction.

September 13, 2026 · Frontier AI

Hassabis Backs a Slower Frontier

Google DeepMind co-founder Demis Hassabis supports Dario Amodei's call to pace AI capability growth so safety work can catch up.

Pacing framework

steps: evaluators, democratic standards, global coordination

Risk window

6–12

months cited for a more capable misaligned swarm

Amodei's ask

unilateral commitment to embedded evaluators

How long is the warning window?

The cited range stretches from six months to one year.

6 months

12 months

The proposal buys time for safeguards; it does not call for a blanket halt.

Why supporters see value

Independent evaluators can verify safety claims.
More time can strengthen alignment and monitoring.

What remains unresolved

Labs must coordinate despite commercial incentives.
Cross-border compliance is hard to verify.

The decision ahead

Can safety systems keep pace with capability gains?

Hassabis's endorsement shifts the debate toward verifiable oversight, thresholds and international coordination.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.