Google DeepMind’s Hassabis backs Huang’s open AI push

Google DeepMind CEO Demis Hassabis has publicly endorsed Nvidia ↘ 0.03% CEO Jensen Huang's push for open-weight artificial intelligence, declaring that "a strong and secure open ecosystem is important for the world to benefit from AI." The statement came days after Huang used his first-ever post on X to publish a coalition letter urging Washington to embrace open AI models.

Huang Breaks Social Media Silence for Open AI Push

On July 24, Huang shared an open letter titled "Open Weights and American AI Leadership," co-signed by more than 20 organizations including Meta ↗ 0.57%, Microsoft ↗ 0.65%, Andreessen Horowitz, Hugging Face, IBM International Business Machines Corporation ↗ 3.96%, Mistral AI, Palantir ↗ 0.83%, Mozilla, and Y Combinator. The letter draws a parallel with the open-source software movement of the 1980s, arguing that America's AI leadership depends on creating an ecosystem where developers, startups, and universities can freely build on powerful models.openthemagazine+3

"AI will transform every industry, power every company, and be built by every country," Huang wrote, adding that "the world needs cutting-edge closed-source models, and it also needs cutting-edge open models".chaincatcher+1

Hassabis Points to DeepMind's Open Science Track Record

In his response on X, Hassabis highlighted Google Alphabet Inc. ↗ 1.53% DeepMind's own contributions to open AI, noting that the Gemma family of open models has now been downloaded more than 300 million times. He also cited AlphaFold, the protein structure prediction system that earned him a Nobel Prize in Chemistry in 2024, as evidence of the benefits of sharing AI breakthroughs openly.x+1

The endorsement is notable given that Hassabis has recently advocated for stronger AI oversight. Earlier this month, he proposed a new U.S.-led AI watchdog modeled on FINRA that would require frontier models — whether open or closed — to pass safety evaluations before deployment.axios

A Shifting Consensus in Silicon Valley

The letter's signatories argue that open-weight models strengthen cybersecurity by allowing researchers to examine systems for vulnerabilities, expand economic access to AI, and prevent excessive concentration of power. The coalition's breadth suggests a growing consensus among major technology leaders that openness and safety need not be mutually exclusive — a position Hassabis's endorsement now reinforces from the research side of the AI industry.openthemagazine+1