Nvidia’s Huang Rejects AI Doomsday Warnings and New Guardrails

Jensen Huang is rejecting both the starkest predictions about AI and the case for new rules around it. In a CBS News interview, Nvidia’s chief executive called claims that AI could end humanity by 2030 “doomsday narratives,” said they were not grounded in science, and declared there was a “0% chance” that 2030 would be the end of the world.

His argument goes beyond dismissing an extinction deadline. Huang said existing U.S. product-liability, cybersecurity and unauthorized-entry laws are sufficient for AI-related harms, and he said he agrees with President Donald Trump that the industry does not need additional guardrails. It is a clear preference for applying current rules over building a separate AI regulatory regime.

The rulebook Huang wants applied

Huang’s formulation treats AI failures as problems that can be addressed through established legal duties. He pointed to liabilities related to cybersecurity and damage, rather than proposing new requirements tailored to AI developers. That does not amount to saying companies have no safety obligation: Huang said Nvidia’s own success depends directly on safe deployment of products and services.

The position arrives as the public argument over AI’s pace has intensified. Huang was responding to recent comments from former Anthropic researcher Jacob Coxon that AI developers earnestly believed the technology could kill everyone by the decade’s end. CBS News also noted calls from Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and other industry figures to slow development.

China policy is a separate point of disagreement

Separately, Huang criticized U.S. limits on Nvidia’s advanced-chip sales to China, while saying technology created in America should benefit the United States first. He said American chip companies should compete worldwide, while identifying national security and defense as areas where U.S. AI companies could not work with China.

He also said he wanted to discuss global standards for AI development with Chinese President Xi Jinping. The interview therefore pairs Huang’s opposition to additional U.S. guardrails with a stated interest in international standards, though he did not specify what those standards would require or who would enforce them.

A concession on data centers

On the infrastructure supporting the AI buildout, Huang made a more concrete concession. He said the industry should have engaged communities sooner, before proposing data centers in places where residents have raised concerns. Huang argued that newer facilities are more energy-efficient, use less water and can support job growth and local economic activity, but his message to communities was an acknowledgment that builders need to show up earlier.