Huang says Nvidia would cut off unsafe AI developers

Nvidia CEO Jensen Huang said he would stop selling chips to a customer that was building unsafe AI. He called the scenario hypothetical and said a conversation would come first. He also said he doesn't think the current level of worry about the technology is justified, Bloomberg reported on Wednesday, October 14.bloomberg.com

Bloomberg's Emily Chang asked Huang in London whether he would cut off an irresponsible developer. "It's a hypothetical, but yes," he said. "But before you do so, you would have a conversation. We don't run other people's companies, and there's a limit to the judgment we apply to other companies' use of our technology. Our technology is general-purpose."qz.com+1

The remark matters because Nvidia leads the market for the accelerators used to train and run AI models, Investing.com noted. Huang did not name a company or say what would count as unsafe.startupfortune.com

Pushback on AI fears

Huang is traveling through Europe ahead of an Nvidia event in Berlin. He described fears about AI safety as growing pains for an industry that is expanding fast, and said he believes the software can be controlled and deployed safely, Quartz reported. He said some of the concern is "self-imposed" and that the industry is doing a "horrible job" of explaining what the technology can do for people.qz.com

He also argued that AI itself can help keep the technology in check. "We should accelerate the development of monitoring technologies, and AI is that monitor. So we should accelerate AI technology," Huang said.ca.investing.com

$250 million for UK safety research

The interview came the same day Nvidia pledged computing resources valued at $250 million over five years for agentic AI safety and security research in Britain, Reuters reported. In its announcement, the company said the deal provides $50 million in compute access each year. Most of it will go to the wider UK research community. Part will go to the Oxford-led BOLD lab, the UCL-led SOFAIR lab and Imperial College London.reuters.com+1

Nvidia said the compute will be delivered through UK-based Nvidia Cloud Partners, and that provider selection, technical architecture and commercial terms have yet to be agreed. The Next Web pointed out that the pledge is compute access rather than cash, and that it will run on Nvidia's own chips through its partners.thenextweb.com

Huang made similar arguments recently in New York. There he said "AI will not run wild" and argued that slowing development for safety's sake would be the wrong choice, SBS reported. "I think we need to accelerate development," he said.news.sbs.co.kr