Nvidia’s Huang calls AI extinction warnings unfounded as the industry splits on slowing down

Nvidia chief executive Jensen Huang told CBS News that fears of AI ending humanity are not grounded on facts¹. He rejected predictions that artificial intelligence could cause human extinction by 2030, putting the chance at 0%², a figure The Verge reported in the same terms³.

Huang went further, saying AI leaders calling for a slowdown are being disingenuous and have ulterior reasons for the warnings, and that current regulations are already enough⁴. TechCrunch framed the week’s exchange as a test of whether the industry is really ready to slow down at all, and singled out Huang’s language as the sharpest contribution to it⁹. The Globe and Mail summarised the split: Huang and Meta’s Mark Zuckerberg say the labs can police themselves, and President Donald Trump has sided with that camp⁸.

Why it matters to investors

The dispute is about capital spending, not philosophy. Nvidia is opposing an industry-wide slowdown in research, and Huang has become a key supporter of the president’s position against one⁵. CNBC reported him emerging as Trump’s top ally in the AI safety debate, a role other technology executives who have courted the administration have not taken⁶.

The commercial logic is visible on the face of it. A coordinated pause on frontier training would reduce demand for exactly the compute Nvidia sells, which is why Huang’s critics read his 0% as a position rather than an estimate^3,4. The counter-argument from the slowdown camp is that self-policing by labs is not a regulatory regime^7,8. Huang’s own position is that the rules already on the books are sufficient, which makes the debate about whether to add anything rather than about what to add⁴.

What to watch

The decisive variable is Washington rather than the labs. With the president aligned against a unified slowdown^5,8, the constraint on frontier development would have to come from Congress, the agencies or the courts rather than from an industry agreement.

The second thing to watch is whether the self-policing position holds as a coalition. It currently rests on Huang and Zuckerberg making the same argument^7,8, with Nvidia carrying the commercial case publicly^5,6. Whether the rest of the industry follows, or the slowdown camp consolidates, is the question the coverage leaves open^1,9. The 2030 horizon Huang was asked about is the date the warnings themselves use, so the disagreement is now anchored to a specific year rather than to an open-ended risk^2,3.