Nvidia CEO Jensen Huang urged AI companies to control their own development pace and withhold products they are not confident are safe, arguing that market forces can support safe innovation. “We don't need any new laws, we don't need any new regulation,” he said at Salesforce's Dreamforce event on Tuesday.

Huang called the choice between fast and safe AI development a “false choice,” saying companies can pursue both. He urged them to move as quickly as possible but pause if their company is out of control or their products are not going to be safe. “The market forces are already there,” he said.

His position contrasts with Anthropic CEO Dario Amodei's call to moderate AI development and introduce outside evaluations and common safety standards. Huang called Anthropic's proposed antitrust waiver “Completely unnecessary.” The waiver would allow Anthropic to coordinate with OpenAI and other AI labs on safety without antitrust concerns.