Nvidia's Jensen Huang says AI needs no new safety regulation

Nvidia founder and CEO Jensen Huang laid out his position on AI safety while speaking at Salesforce’s Dreamforce conference on Tuesday. Rejecting the idea — voiced by at least one OpenAI safety researcher — that AI amounts to some kind of “alien mind,” Huang argued it’s simply hardware and software built by humans, and can therefore be controlled by humans and existing laws. “Safety is an engineering problem, not a legal one,” he said.

Huang sees little need for new AI laws or regulations at all, arguing that free-market pressure will be enough to keep companies from releasing unsafe products. Companies that aren’t confident in a product’s safety, he said, simply shouldn’t release it, and should pace themselves until they are. He said treating innovation, speed, and safety as mutually exclusive was a false choice, and that a company can run fast while still pausing if it feels something is out of control.

Huang has been building the hardware behind AI since before ChatGPT existed, and Nvidia now also makes open-source models, agents, harnesses, and sandboxes — though skeptics might note his own company has a business stake in AI moving as fast as possible. Even well-intentioned companies ship faulty products: the 2024 CrowdStrike outage grounded thousands of flights, and Meta recently paid $18 billion to settle a lawsuit over social media’s harms to children. AI has caused its own harm too, from an OpenAI model hacking into Hugging Face to lawsuits against an AI lab over the suicides of young people who had long conversations with its chatbot.

Huang didn’t address industry self-regulation, an approach that appears close to taking shape and which he has pushed via open-weight models as a counterweight to proprietary AI labs. Microsoft CEO Satya Nadella, speaking at the All-In Summit on Monday, argued that China should care about the same AI safety concerns as the U.S., since it faces the same hacking risks and also wants its citizens to benefit from AI. Huang, for his part, demonstrated this week that he has direct access to President Trump.

Why it matters

If Huang’s “leave it to us” approach prevails, the industry — not lawmakers — would set the pace on AI safety even as products from major labs have already caused real-world harm, from data breaches to lawsuits over user deaths.