Nvidia CEO Jensen Huang and Meta CEO Mark Zuckerberg both rejected Dario Amodei's call for AI labs to slow down this week, telling the industry's most prominent safety advocate that a coordinated isn't happening.
On Monday at the All-In Summit in Los Angeles, President Trump called into Jensen Huang's onstage interview just as the conversation turned to Amodei's proposal. Trump warned that anyone pushing a slowdown was "playing right into the hands" of political rivals, or China. Huang's answer came fast. "We're not going to let that happen, sir," he said, according to TechCrunch, and the crowd applauded.
That exchange was a direct answer to an essay Amodei published days earlier called "We Must Pace the Frontier." In it, the Anthropic CEO argued the industry should deliberately slow how fast it improves model capabilities. He went further, too: he asked Washington to issue a narrow antitrust waiver so rival labs could hold safety conversations without breaking collusion law. Truth on the Market put it bluntly: competitors agreeing on how fast to build is, legally speaking, a cartel. Sam Altman endorsed the idea within days. So did Elon Musk.
Huang didn't reject all of it. One piece of Amodei's proposal he agreed with: letting independent third parties run safety evaluations at frontier labs. Pacing, though, he rejected outright. A coordinated slowdown, he argued, would hand an edge to countries racing to catch up with US AI.
A day later, Zuckerberg made a similar case at Dreamforce in San Francisco. "Trust and alignment are quickly becoming the most important capabilities that will differentiate agents and models," he said. His argument: labs already have enough incentive to build safely. A company whose model causes real harm faces what he called significant liability. No cartel required.
[Hundreds Banged Pots Outside Sam Altman's G20 AI Summit in Chapel Hill](https://startupfortune.com/hundreds-banged-pots-outside-sam-altmans-g20-ai-summit-in-chapel-hill/)
While Sam Altman, Jensen Huang and Alex Karp pitched AI as essential infrastructure at the G20 Innovation Ministerial in Chapel Hill, hundreds of protesters banged pots and pans outside the Carolina Inn to mimic data center noise. The demonstration tied together job-loss fears, UNC's Flock surveillance cameras and a wave of over 30 North Carolina... - [protesters banged pots outside Sam Altman's AI summit](https://startupfortune.com/hundreds-banged-pots-outside-sam-altmans-g20-ai-summit-in-chapel-hill/) - [G20 AI summit Chapel Hill data center noise](https://startupfortune.com/hundreds-banged-pots-outside-sam-altmans-g20-ai-summit-in-chapel-hill/)
He got specific, too. Zuckerberg said Meta Superintelligence Labs already uses independent evaluators "in several areas" and called that industry best practice, then added that "other labs can just do this too" instead of waiting for regulators to bless a shared framework. He pointed to Meta's own record as proof the market already works: the company delayed shipping its Muse AI models over safety and security concerns, he said, with no antitrust exemption needed to do it. "We just did it as part of our day-to-day work because it was clearly the right thing for people and for us."
Amodei's ask is bigger than it sounds. Rival AI labs agreeing to limit how fast they ship new capabilities is exactly the kind of deal antitrust law exists to block, so Amodei wants Washington to carve out an exception, mediating or at least enabling safety talks between competitors that would otherwise be illegal collusion. That is a real regulatory lift, not a talking point, and it needs Congress or federal agencies to move. It also needs the industry's biggest players to actually want it.
They don't, not all of them. Huang and Zuckerberg just told Amodei so, in public, on back to back days.
Two camps, heading into 2027 #
The lineup is no longer subtle. Anthropic and OpenAI, closest to whatever model generation comes after Claude's current line and GPT-5, are the labs asking for a pace. Nvidia and Meta, whose businesses depend on more compute getting built and more models shipping faster, are refusing. Trump made his own view clear on the same call with Huang, calling AI existential risk a "hoax" and adding that "the robots are not going to be taking over the world," remarks reported by Fox Business and Yahoo News.
Frankly, an antitrust waiver that Nvidia and Meta won't use is a plan two labs can follow while the rest of the industry keeps racing. Amodei wanted coordination. What he got, so far, is a split.
Also read: As AI-Written Content Floods The Web, Lynote's AI Humanizer Bets Businesses Still Want To Sound Human • The Fed Raises Rates to 4% Anyway, Defying Trump's Pressure Campaign • Ryan Greenblatt Built a Hotline for AI Agents to Snitch on Each Other
Nvidia CEO Jensen Huang has publicly criticized existential AI risk warnings from leaders including Anthropic's Dario Amodei, arguing that apocalyptic framing reflects a God complex and damages the industry's ability to attract workers and investment. The dispute is not merely philosophical: the competing narratives from Huang and Amodei are... - how to price a SaaS product for enterprise - cold email template that gets replies from investors
This article is posted in AI News, check it out for more related stories.
Join the discussion #
Open in the community → Almost there. Sign in and your reply posts straight away.