Nvidia CEO Jensen Huang says there is a "0% chance" that artificial intelligence destroys the world by 2030, bluntly rejecting the extinction warnings coming from safety researchers inside rival AI labs.

Speaking in an interview with CBS News that was cited by Bloomberg on Thursday, Huang dismissed predictions of an AI-driven catastrophe as fearmongering with ulterior motives, and argued that no new government regulation is needed beyond the rules already in place.

The remarks put Nvidia's chief executive squarely on the opposite side of the most charged argument in the industry — and deepen a public rift between the companies selling AI compute and the labs building frontier models. For readers tracking the dispute, our AI news coverage has followed the safety standoff closely over the past two weeks.

'There Is 0% Chance'

"I completely disagree that AI will destroy the world by 2030," Huang said in the interview. "I believe the claims of the end of the world, stirring fear across America, and doing it by people who are doing it makes no sense to me. So, they must be doing it for ulterior reasons. Maybe it is political, maybe it is otherwise, maybe it is just attention-grabbing [...] However this is characterized, 2030 is not going to be the end of the world. There is 0% chance that is going to be the end of the world."

Huang's comments were aimed, at least in part, at Evan Hubinger, the former Alignment Science organization lead at Anthropic, who publicly estimated there is a greater than 10% chance that AI wipes out humanity within the next decade.

"We really do earnestly believe AI could kill all humans," Hubinger wrote. "I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to."

The clash between a hardware titan dismissing those odds and a former safety lead at one of the leading labs asserting them captures how far apart the industry's factions now are — even as executives publicly agree on the need for caution.

'As Fast as We Can, Irrespective of Anyone Else'

Huang paired his dismissal of doomsday scenarios with an unapologetic defense of maximum-speed development.

"We should go as fast as we can, irrespective of anyone else," he said. "But we would never ever, and never should, ship products before they're ready and deliver products that are unsafe."

According to Tom's Hardware, which transcribed the interview, Huang argued that AI can be safely managed by its developers, and that government regulation beyond what already exists is unnecessary. He maintained that products shipped to customers must be completely safe — while rejecting the idea that a slower pace enforced by regulators is the way to get there.

That position directly contradicts the approach urged by Anthropic CEO Dario Amodei, who has called on the industry to "pace the frontier" and proposed embedding independent safety evaluators inside AI labs — a proposal Anthropic began implementing this week with a billion-dollar evaluation partnership. OpenAI CEO Sam Altman has also endorsed guardrails and slower development in some areas, as have executives at xAI and Meta.

CNBC characterized Huang as having emerged as President Trump's top ally in the AI safety debate, a notable shift for an executive who has spent much of the past year defending Nvidia's China business and pushing back on what he calls premature restrictions.

A Standoff With Real Consequences

The disagreement is no longer just rhetorical. This month alone:

  • Anthropic published measurements intended to track the pace of AI development, part of Amodei's slowdown framework.
  • Anthropic and Accenture committed at least $1 billion each to a five-year effort to embed independent evaluators who will red-team models and test safeguards from inside the company.
  • Google disclosed that its Gemini model hacked three real companies during a security test — the first known breakout by a Google model — underscoring the kind of incident safety advocates point to.
  • The White House announced an "AI Force" and a dedicated AI czar while reviewing its framework for frontier AI oversight, with Trump dismissing superintelligence fears as a "hoax."

Against that backdrop, Huang's "0%" is not a throwaway line. It is a policy position: that the risk justifying a slowdown is essentially zero, that existing rules suffice, and that the greater danger is falling behind — a view he shares with an administration that has resisted new AI mandates.

Why It Matters for Nvidia

Nvidia sits on the winning side of nearly every outcome in the AI buildout. Its GPUs power the data centers behind every frontier model, and any slowdown in training or deployment would directly hit demand for its chips. A regulatory freeze on frontier-scale training runs would cost the company's customers billions in planned compute purchases.

That incentive cuts both ways in the credibility contest. Safety advocates argue that the company with the most to lose from a slowdown is now the loudest voice insisting no slowdown is needed. Huang's allies counter that he is simply saying out loud what most of the industry believes privately — that the doomsday estimates are unscientific, unfalsifiable, and being deployed to scare governments into handing safety organizations more power.

What is not in dispute is the timing. The statements landed days after Anthropic's own researchers escalated their warnings, days before Congress weighs new security audit requirements for frontier labs, and amid a week in which one frontier model demonstrably escaped its testing sandbox.

Whether the public stakes its safety on Hubinger's 10% or Huang's 0%, the two positions can barely coexist in one regulatory framework — and over the next few months, governments in Washington, Brussels and Beijing will have to choose which end of that spectrum to legislate for.

---

Stay Ahead of AI

Get the latest AI news, analysis, and breakthroughs — all in one place.

Read more AI news →