Nvidia CEO says there’s ‘0% chance’ AI wipes out humanity by 2030, despite

Nvidia CEO Jensen Huang has dismissed the idea that artificial intelligence could lead to human extinction by the end of this decade, according to CBS. He said there is a 0% chance of such an outcome by 2030 and described the most apocalyptic predictions as unsupported by science. However, Huang made it clear that he does not view AI as entirely free of risk, and he said development can never come at the expense of safety.

Huang is specifically arguing that the scenario where AI wipes out humanity by 2030 is not a credible one. This puts him in a different camp than some former and current researchers who have raised alarms about the pace of development. The debate gained significant traction after Jacob Coxon, a former researcher at both Anthropic and OpenAI, resigned and published a series of warnings.

Coxon argued that AI companies are racing toward self-improving superintelligence and suggested that people working inside these labs genuinely believe the technology could be lethal to humanity before the decade ends. Evan Hubinger, who works on AI alignment at Anthropic, has stated he believes there is a greater than 10% probability of human extinction from AI within the next decade.

Some of the people building AI think the odds are far worse

Another Anthropic researcher, Samuel Marks, said senior researchers frequently take the possibility of extreme AI danger seriously. These figures are individual assessments rather than empirically measured probabilities, as no scientific model can currently establish whether the true risk of extinction is 0%, 5%, 10%, or any other specific number.

The industry is clearly divided on how to manage these risks. Anthropic CEO Dario Amodei has called for a slower pace of frontier development until stronger safeguards are in place, as he describes the rate of AI improvement as exponential. Similarly, OpenAI CEO Sam Altman has agreed that frontier development should be paced rather than recklessly accelerated.

Huang disagrees with the idea of slowing down the industry or creating a large new regulatory regime. He argues that AI companies already face existing laws covering product liability and cybersecurity. His preference is to develop AI as fast as possible without compromising safety, engineering it directly into the products.

Huang even suggested that those promoting extreme danger narratives might have ulterior motives, ranging from political goals to a desire to attract attention. He also said companies pushing for new AI-specific rules might be trying to avoid responsibilities they already have under existing law.

Anthropic, which walked back a core safety commitment earlier this year, has told the U.S. government it expects systems matching or exceeding top human experts as early as late 2026 or early 2027. However, researchers such as Gary Marcus, Melanie Mitchell, and Yann LeCun remain skeptical that present-day neural networks show uncontrollable superintelligence is imminent.

Nvidia, valued at roughly $5 trillion, sells the GPUs and computing systems that power the AI boom. The company most recently reported quarterly revenue of $96.2 billion, a 106% increase year over year. Huang told the outlet that Nvidia’s success is directly tied to the safe deployment of AI, as its value would suffer otherwise.

Huang is due at the White House on Thursday for a dinner with Chinese President Xi Jinping, where he said he wants to discuss global standards for AI development.