Nvidia CEO Jensen Huang: We would stop supplying customers who develop unsafe AI

Nvidia CEO Jensen Huang said in a Bloomberg interview in London that the company would stop supplying chips to customers found to be developing unsafe AI software, though it would talk to them before cutting off supply. He emphasized that Nvidia's technology is general-purpose and the company has limits in judging how others use it.

Huang downplayed concerns about AI safety, saying fears had been overstated and that talk of an AI bubble was exaggerated. On Oct. 14, as Nvidia and other chip stocks fell, he repeated his optimistic view on AI infrastructure demand, saying supply-demand imbalance would last for years. "Supply and demand will eventually balance, but certainly not today," he said at the London event. He also pushed back on calls for a slowdown across the industry and described some worries as "self-imposed," adding that the industry has done a poor job explaining technology's benefits to the public.

Huang proposed using AI to monitor and govern AI, saying the industry should accelerate the integration of such safeguards into AI systems. Nvidia also announced a $250 million commitment over five years to UK AI safety research, with funding provided in the form of advanced computing power. Institutions including Oxford University and University College London will participate, with research covering AI safety protections and methods for governing AI agents.