Jensen Huang backs industry-led AI safety decisions

Speaking at a Salesforce conference in San Francisco on Tuesday, Huang said there should be no “false choice” between the speed of AI innovation and the safety of AI products.

“We don’t need new laws or regulations,” Huang said. He argued that AI companies should essentially regulate themselves and decide not to release products when their leaders lack confidence in their safety.

“Safety is the most important thing. But safety is an engineering issue,” Huang said.

He added that companies should pause development or release plans if they believe their AI systems are out of control.

“That makes sense,” Huang said. “Escape as quickly as possible, but if you feel the facility is out of control, pause.”

AI safety concerns are growing

Nvidia is the world’s largest company by valuation, and its profits have soared because of strong demand for the AI computing chips it makes.

Huang’s position on AI safety contrasts with concerns recently expressed by industry insiders who warn that advanced AI could pose an existential threat to humanity.

Last week, a post by an artificial intelligence researcher who left Anthropic went viral online. The researcher claimed that, if left unchecked, AI could kill all of humanity by the end of 2010.

Other AI executives and experts said they agreed. Over the weekend, Anthropic CEO Dario Amodei called for slowing the pace of all AI development, external and urged the government to regulate the industry.

The post was praised by OpenAI chief executive Sam Altman, Google DeepMind co-founder Demis Hassabis and Elon Musk, owner of the social media platform X and AI assistant Grok.

However, some industry insiders say the concerns are exaggerated and may be contributing to hype around the AI industry.

Critics oppose relying entirely on AI companies

Some people in the industry believe allowing AI executives to regulate themselves is a bad idea.

Anthropic executive and co-founder Jack Clarke told the BBC on Monday that leaving AI as a “totally unregulated industry” would be “a roll of the dice with enormous risks”.

Patrick Hillman, CEO of Logical Intelligence, whose chairman is AI industry expert Yann LeCun, also criticized the public’s lack of trust in technology companies. He questioned how little faith people have in technology companies to act in the public interest, external.

“The only institution Americans may trust less than Washington these days is Silicon Valley. I’ve worked and lived in both, and I’m sure you share this skepticism,” Hillman said.

“If you think what you’re making is dangerous, show us you’re willing to stop doing it,” he added.