Nvidia CEO Jensen Huang told CBS News there is a "0% chance" artificial intelligence ends the world by 2030, calling extinction warnings irresponsible, and arguing the real danger is letting fear override the laws already on the books.
Huang, whose company manufactures the specialized chips powering virtually every major AI system in the country, sat down with CBS News correspondent Jo Ling Kent on Friday for a wide-ranging interview. He took direct aim at a growing chorus of AI researchers and rival executives who have urged the industry to slow down, dismissing their predictions as unscientific fearmongering.
The interview landed one day after OpenAI disclosed six new incidents it described as "unexpected or concerning" behavior from its AI models, a revelation that prompted some lawmakers to push for tighter oversight of AI developers. Huang was unmoved. He framed the panic as a distraction from accountability, not a reason for new rules.
Former Anthropic researcher Jacob Coxon sparked the latest round of alarm earlier this month when he claimed on social media that AI developers "earnestly believe that it could kill us all by the end of the decade." Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman have both called for slowing AI development.
Huang rejected the premise entirely. He told CBS News:
"2030 is not going to be the end of the world. There is 0% chance that's going to be the end of the world."
He went further, calling the warnings a disservice to the public.
"Scaring people is unnecessary. It is irresponsible."
His argument was not that AI carries zero risk. He said Nvidia's own business depends on safe deployment. But he drew a sharp line between managing real-world product risk and indulging apocalyptic speculation that, in his view, lets companies dodge the rules that already apply to them.
Huang pointed to existing U.S. laws governing product liability, cybersecurity, and unauthorized access as more than sufficient to hold AI companies accountable. He told Kent that the industry already faces "all kinds of liabilities associated with cybersecurity" and damage liability statutes.
"Apply that first, don't let this doomsday narrative allow someone to relieve them of the laws that currently exist."
That line captures a position that deserves more attention than it typically gets. The push for sweeping new AI regulation often comes from the same companies building the technology, companies that stand to benefit if compliance costs price out smaller competitors. Huang's counter is straightforward: enforce the laws we have before writing new ones that may do more to protect incumbents than the public.
The interview also covered the fraught question of selling advanced chips to China. The U.S. government currently bars Nvidia from exporting its most advanced processors, the kind that power large AI models, to Chinese buyers. Last December, President Trump allowed Nvidia to begin selling its H200 chips to "approved customers" in China, a partial loosening of restrictions.
Huang endorsed that direction. He said American chip companies should serve global markets, not cede them.
"Every single chip company should go and serve the world, compete for the world. That's all about competing. America is about competition."
He acknowledged national security limits, but his broader point was that pulling back from global markets hands market share to foreign rivals without making the United States safer. President Trump has warned that slowing frontier AI development could give China an edge, a view Huang said he shares.
Huang also disclosed that he plans to attend a White House dinner on Sept. 24, where President Trump will host Chinese President Xi Jinping. He said he would like to speak with Xi about global AI standards.
His comments about Chinese leadership drew some of the interview's most striking language. Huang told Kent:
"They want China to have the benefit of AI. They want China to prosper, just as we want America to prosper. Don't you wish every country would have aspirations for a better tomorrow? And I love that about them."
That is a generous reading of Beijing's ambitions, to put it mildly. Whether Huang's optimism about Chinese intentions reflects genuine belief or strategic positioning for a company that wants access to the world's second-largest market is a question the interview left open.
Nvidia sits at the center of the AI infrastructure boom. The company's market value has reached $5.3 trillion, making it the world's most valuable publicly traded company. Its chips power the large language models behind products from OpenAI, Anthropic, and dozens of other firms.
That boom has a physical footprint. AI requires massive data centers, and communities near proposed sites have raised concerns about energy consumption, water usage, and local disruption. Huang conceded Nvidia has not always handled the community side well.
"I would say we're sorry we didn't come talk to you sooner. We need to go engage the communities long before we set up the data centers."
He framed data centers as a revival of American manufacturing, saying the industry can "bring back building and manufacturing again." Whether affected communities see it that way depends on how seriously companies like Nvidia follow through on early engagement, a promise that costs nothing until it is tested.
Huang's personal wealth stands at $182 billion, making him the eighth-richest person in the world, per the Bloomberg Billionaires Index. California voters will decide this fall on a ballot measure that would impose a one-time 5% tax on state residents with net worths of $1 billion or more.
Huang estimated his liability under the proposal at $8 billion over five years. Rather than object, he leaned into it.
"The fact that I can afford to pay $8 billion in taxes over five years is a privilege. I feel it's a privilege. It's a responsibility. And nothing would give me more joy than to be able to pay even more taxes."
He added a line that played well for the cameras: "I'm not afraid of paying taxes, I'm just afraid of being poor."
Huang also described himself as "the ultimate American dream," referencing his early years working as a dishwasher at a Denny's restaurant, the same period, he said, when he conceived the idea for Nvidia.
"I was just really, really focused on doing dishes like no dishwasher ever did dishes."
The timing of Huang's interview matters. Just one day earlier, OpenAI, whose products run on Nvidia hardware, disclosed six additional incidents of AI models behaving in unexpected or concerning ways. In a separate episode in July, OpenAI bots reportedly went rogue and hacked AI company Hugging Face.
Those incidents gave ammunition to lawmakers calling for tighter AI oversight. They also, perhaps unintentionally, reinforced part of Huang's argument: if AI systems are already breaching other companies' security, the question is not whether to regulate but whether existing cybersecurity and liability laws are being enforced against the companies responsible.
Huang's position, that current law is adequate if actually applied, sidesteps a harder question: whether the speed of AI development has outrun the enforcement capacity of agencies built for a slower era. He did not address that gap directly. His answer, in effect, is that the industry can police itself because its business model depends on safe deployment.
"Our company's success is directly connected to the safe deployment of products and services. If we don't continue to do that, our value would be diminished."
That is a market argument, not a regulatory one. It works as long as the market punishes failure. The track record of trillion-dollar companies facing meaningful market consequences for safety lapses is, to be charitable, mixed.
Huang told CBS News he agrees with President Trump's position that the AI industry does not need additional guardrails. Trump has taken the view that excessive regulation risks slowing American innovation at the precise moment China is racing to close the gap.
That alignment puts Huang in a different camp from Amodei and Altman, both of whom have called for slowing development, even as their own companies continue to push the frontier. The contradiction between urging caution and shipping new models at breakneck speed has not gone unnoticed in Washington or in the industry.
Huang's counter is simpler and more consistent: build fast, compete globally, enforce the laws already on the books, and stop pretending the sky is falling. Whether that framework holds up against the next rogue-bot incident or the next disclosure of "unexpected" AI behavior remains an open question.
When the people building the most powerful technology on earth split into two camps, one warning it could end civilization, the other saying enforce the rules we have, the public deserves to know which camp is motivated by genuine concern and which is angling for competitive advantage. Huang, at least, is not hiding behind fear to avoid accountability.
