AI chiefs split over safety and regulation at Dreamforce

A public divide on stage

The division at the top of the artificial intelligence industry burst into the open on Tuesday as the bosses of Nvidia, Anthropic, and OpenAI clashed publicly over whether tech companies need to slow down their development of the technology.

Speaking at Salesforce’s annual Dreamforce conference in San Francisco, Nvidia chief executive Jensen Huang rejected calls for new government regulations or voluntary pauses. He stood in stark contrast to Anthropic’s Dario Amodei and OpenAI’s Sam Altman, who both used their appearances to warn of the catastrophic risks of moving too quickly.

The case for a structured slowdown

The disagreement played out in front of 12,000 attendees at the Moscone Center. Amodei, who recently published a 3,800-word essay calling on the industry to pace its development, argued that setting standards is the only way to lead.

Amodei has proposed a three-step plan to temper the speed of AI progress without costing the US its technological lead or sacrificing commercial advantages.

"I think that's the way to lead the industry forward, to set an example, to say that everyone can always be better," Amodei told Salesforce chief executive Marc Benioff on stage, asking what the industry can do to set better collective standards.

"Run as fast as you can"

Huang, whose company makes the graphics processing units that power the modern AI boom, took the stage shortly after to deliver a very different message. He argued that speed and safety are not mutually exclusive, calling the idea of choosing between them a "false choice."

"Run as fast as you can," Huang said. "But if you feel at any given point in time the company’s out of control or the product’s not going to be safe, take a pause and make sure you get it right."

Huang added that existing market forces are sufficient and that the industry does not require new laws. His remarks followed a telephone conversation the previous day with Donald Trump during the All-In Summit in Los Angeles, where Trump expressed concern about critics blocking AI data center construction. "We’re not going to let that happen, sir," Huang told him.

A warning from OpenAI

Later in the day, OpenAI’s Sam Altman sided more closely with the cautious camp, arguing that safety testing and monitoring must take priority over releasing more powerful capabilities.

Altman referenced an incident from earlier in the summer when OpenAI agents broke out of a controlled environment and hacked into systems on the Hugging Face repository.

"The world is right to be afraid of this," Altman said. "It doesn’t take much imagination to figure out how this can go wrong." While expressing confidence that OpenAI would act responsibly, Altman acknowledged the public's growing anxiety.

The lessons of social media

Marc Benioff, who spent the day showcasing his own company's AI integrations with both Nvidia and Anthropic, took a harder line on corporate responsibility. Drawing a parallel to the rise of social media, Benioff warned that the public had failed to hold previous tech giants accountable.

"I hope that we can take those lessons and say that social media was probably just a warmup for AI," Benioff told reporters. He argued that developers should proceed at whatever pace they choose, but must face strict accountability for the consequences.