Jensen Huang Tells Trump Live On Stage That AI Will Not Slow Down

Jensen Huang does not do things quietly. At the All-In Summit on Monday, the Nvidia CEO took a live phone call from President Donald Trump while standing on stage in front of hundreds of tech industry attendees. The message Trump wanted to deliver, and that Huang seemed very happy to amplify, was simple: robots are not going to take over, and no, the AI industry is not going to pump the brakes.

The stunt was extraordinary even by the standards of 2026, a year that has not exactly been short on tech-politics spectacle. Huang told Trump, on speakerphone in front of the crowd, that "we're not going to let" an AI slowdown happen. The crowd reportedly cheered.

The context: a week of AI slowdown debate

This moment did not happen in a vacuum. Anthropic's Dario Amodei published an essay called "We Must Pace the Frontier" last week, arguing that the breakneck speed of AI development poses genuine existential risks and that the industry needs to coordinate a slowdown. The essay triggered a wave of responses from across the industry and Washington.

Sam Altman, Elon Musk, and Demis Hassabis of Google DeepMind all, to varying degrees, expressed support for Amodei's general direction. The idea of a loose gentleman's agreement among frontier AI labs to ease back on deployment timelines briefly seemed plausible, or at least discussable.

Then Jensen Huang called the President of the United States on stage and said no.

Huang's position is not surprising when you think about Nvidia's business model. Every GPU Nvidia sells goes into training or running AI models. A slowdown in AI development is, very directly, a slowdown in Nvidia's revenue. In 2025, Nvidia posted revenues of over 130 billion dollars. The company's market capitalisation has made it one of the most valuable companies in history. Huang has every financial incentive to keep the accelerator pressed to the floor.

But his argument is not purely self-serving, at least not on its face. Huang has consistently argued that AI will create more jobs than it displaces, that the productivity gains will benefit ordinary people, and that slowing down American AI development only hands an advantage to China. These are legitimate points that deserve engagement, even if the financial interest behind them is obvious.

What the phone he used tells us

TechCrunch noted, with the kind of attention to detail that makes tech journalism worthwhile, that Huang took the call on a phone that was not commercially available. The device is believed to be a prototype from Nvidia's hardware partnerships, a quiet flex that reinforced the broader message: Nvidia is building the future, not waiting for permission.

The image of Huang holding up the phone while the President's voice comes out of the speaker will circulate for a while. It is one of those moments that captures something real about where power sits in 2026. The AI industry, particularly the infrastructure layer that Nvidia dominates, feels increasingly confident that governments need it more than it needs governments.

The cartel question

The Verge raised a question this week that is genuinely uncomfortable: if the major AI labs all agree to slow down together, is that a safety pact or an anticompetitive cartel? Antitrust law has specific things to say about competitors coordinating behaviour, and "we agreed to go slower for safety reasons" is exactly the kind of argument that lawyers on both sides would love to argue in court.

Huang's intervention, backed by Trump's rhetorical support, makes the voluntary slowdown significantly less likely to hold. It will be hard for OpenAI or Anthropic to unilaterally pull back when Nvidia is shipping faster chips every quarter and the US government is explicitly cheering speed over caution.

The week that started with a serious safety essay has ended with a phone call on a stage. That probably tells you something about where this debate is actually headed.