Demis Hassabis Backs Dario Amodei’s Call To Slow Down Frontier AI Race, Warns of Escalating Risks

Google DeepMind co-founder Demis Hassabis has thrown his support behind Anthropic CEO Dario Amodei's proposal for an industry-wide effort to slow the pace of advanced artificial intelligence development, in a move that has drawn immediate backing from figures across the tech and political sectors. Amodei's call to 'pace the frontier' comes amid growing anxiety within the industry over the possibility of uncontrollable AI systems, prompting tech leaders and policymakers to push for formal governance and third-party oversight before models outpace human supervision.

Tech Leaders Rally Around Pacing Development

Hassabis, who serves as chair of Google DeepMind and chief scientist at Alphabet, endorsed Amodei's essay in a post on X, calling it a necessary framework for managing escalating AI risks. He said the direction of the proposal was correct for the moment, even as the details still need to be worked through, and noted that it aligns with Google DeepMind's own recent push to build an industry-wide standards body focused on frontier AI safety. Google DeepMind Researcher Josh Engels Quits AGI Safety Team To Join METR Amid AI Risk Warnings.

Google DeepMind's Demis Hassabis Backs Anthropic CEO Dario Amodei's Call To Slow Down Frontier AI Development

Amodei has argued that while advanced AI carries the potential to cure major diseases and accelerate global economic growth, unchecked commercial competition risks triggering a race to the bottom among AI developers. He warned that without calibrated limits, advanced models could introduce catastrophic hazards, including large-scale cyberattacks, bioterrorism misuse and recursive self-improvement that slips beyond human control. Elon Musk, Sam Altman, Rishi Sunak and Others React to Anthropic CEO Dario Amodei Call for Slowing Down AI Frontier Models.

Anthropic Commits To Embedded Oversight

To kick off the effort, Amodei laid out a three-part framework meant to establish a balanced and verifiable pace for frontier AI development. The plan calls for frontier AI firms to grant continuous, employee-level access to external evaluation teams so they can inspect training and alignment protocols, for developers to coordinate on standardised thresholds that prevent unchecked acceleration, and for governments and international institutions to establish aligned safety baselines across borders. Anthropic has already adopted the first step, committing to give independent third-party evaluators permanent access to its internal systems to assess model behaviour during training and verify adherence to safety standards.

Political Leaders And Advisers Push For Government Action

The proposal has drawn reactions beyond AI labs as well. Former UK Prime Minister Rishi Sunak, who advises Anthropic, said voluntary self-regulation is no longer sufficient given the rapid capability jumps seen in recent models, arguing that governments need to step in with statutory frameworks similar to those governing nuclear energy and advanced biology. He said that while strategic competition between nations remains a factor, major powers must hold direct talks to avert systemic catastrophes.

In the United States, Senator Bernie Sanders went further, describing a development slowdown as only a preliminary step and calling for a comprehensive pause on advanced frontier systems along with an outright ban on autonomous artificial superintelligence. He urged international negotiations between global leaders to codify binding safeguards.

Internal Unease Drives Public Debate

The coordinated calls for a slowdown come amid visible friction inside frontier research labs, with recent high-profile resignations at Anthropic and other major companies spotlighting internal unease over whether commercial pressures are overriding safety protocols. With leading figures at Anthropic, Google DeepMind and OpenAI acknowledging extinction risks and safety gaps, the industry is facing mounting pressure to move from voluntary corporate commitments toward enforceable, internationally verified standards.

(The above story first appeared on LatestLY on Sep 14, 2026 08:42 AM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).