Huang: the labs want to be 'relieved of the laws we do have'
Nvidia's chief executive spent Sunday telling CBS that the industry's own doomsday warnings are a lobbying tactic — and he named the ask he thinks sits behind them.
Jensen Huang says the AI leaders warning about extinction are not asking for new rules — they want out of the ones already on the books. In an interview aired Sunday on CBS News, the Nvidia CEO dismissed the prediction that AI could end the world by 2030 as "not grounded in science," put the odds at "0%," and called the rhetoric "irresponsible" and "unnecessary." Pressed on why the labs would make claims he considers baseless, he said only: "They must be doing it for ulterior reasons."
The sharper part of his answer was about liability, not capability. Existing cybersecurity and damage-liability law already covers the harms the labs describe, Huang argued: "Apply that first — don't let this doomsday narrative allow someone to relieve them of the laws that currently exist." That reframes the pacing debate the industry has been having all month. Anthropic's Dario Amodei and OpenAI's Sam Altman publicly asked for a slowdown in frontier development, and the second phase of Amodei's proposal asked Washington for an antitrust waiver so the leading labs could coordinate release schedules without tripping cartel law. Huang's read is that the safety framing is how incumbents freeze a market they are currently winning.
Nvidia has the least disinterested position of anyone in the argument — a $5.3 trillion market value built on the same buildout the labs want to pace. Huang addressed that directly: "Our company's success is directly connected to the safe deployment of products and services," he said. "If we don't continue to do that, our value would be diminished." It is a fair point and a weak control: no liability regime in any industry has ever been satisfied by a vendor promising that safety is good for the share price.
What makes this fight genuinely hard to referee is that both sides profit from the same alarm. The labs get a regulatory moat and a reason to stop burning capital on pre-training runs with diminishing returns; Nvidia gets to sell compute into a market with no speed limit and no liability exposure. Huang is not the neutral party here — he is the other interested party, and he is the only one willing to say so out loud.
He also put the argument in a geopolitical frame. Huang said he will attend a White House dinner on September 24, where President Trump hosts Xi Jinping, and that he wants to raise global standards for AI development with the Chinese leader. He opposes US bans on selling advanced chips into China on the grounds that American firms should compete there, while conceding national security and defense work is a separate category. On the domestic backlash against data centers, he offered an apology of sorts to the communities hosting them: "We're sorry we didn't come talk to you sooner."
We covered the earlier round of this fight — Trump's answer to the pacing ask: he already has criminal power over the labs.
What to watch: whether the September 24 dinner moves export rules, and whether any frontier lab answers the "ulterior reasons" charge on the record rather than through a spokesperson.
If the labs' warnings really are a liability play, what would it take for you to believe them?
