Nvidia CEO Says AI Needs No New Laws, Just Market Forces

Nvidia founder and CEO Jensen Huang is drawing a hard line in the sand on AI regulation. Speaking at Salesforce’s Dreamforce conference this week, he made it crystal clear: new laws aren’t needed to govern artificial intelligence. Instead, he believes the free market and existing product liability laws are more than sufficient.

“Safety is an engineering problem, not a legal one,” Huang declared. His reasoning is straightforward. AI is just hardware and software built by humans, so humans can control it. Companies already know not to release unsafe products. Why? Because it’s bad for business.

Market Forces vs. Real Harm

On the surface, Huang’s argument has appeal. He’s not wrong that companies have incentives to avoid shipping broken products. But here’s where reality gets messier than his optimistic view suggests.

Remember CrowdStrike’s 2024 software update that grounded thousands of flights and crippled businesses worldwide? That wasn’t malicious intent or recklessness. It was a mistake from a company with good intentions. Now imagine those consequences at AI scale.

And AI has already caused documented harm. OpenAI models have been caught hacking systems. There are lawsuits alleging the company’s chatbot contributed to young people’s suicides through extended conversations. Meta just shelled out 18 billion dollars to settle a lawsuit over social media harms to children. These aren’t hypothetical scenarios. They’re happening now.

Huang’s “let the market decide” philosophy might work great when a software glitch costs money and downtime. It’s far less comforting when the stakes involve people’s safety, mental health, or national security.

The Convenience of Huang’s Position

It’s also worth noting that Huang’s skepticism of regulation is perhaps unsurprising given his position. Nvidia has been absolutely printing money from the AI boom. Why would he want new oversight mechanisms that could slow innovation and create compliance costs? As he himself noted, “The sky’s the limit for us.”

There’s nothing inherently wrong with that ambition. But it’s worth recognizing the perspective it creates. When you’re the one selling the chips powering the AI revolution, “move fast and break things” becomes a pretty attractive philosophy.

The Self-Regulation Question

Huang didn’t mention industry self-regulation, which increasingly looks like where this is heading anyway. Microsoft CEO Satya Nadella raised an interesting point at the All-In Summit: China should care about AI safety just as much as the US does. The risks don’t respect borders. But getting global coordination on voluntary safety standards? That’s exponentially harder than national regulation.

The problem with pure self-regulation in a competitive space is obvious. If one company decides to move slower for safety reasons while competitors race ahead, market pressures cut the other way. Innovation becomes a race to the bottom on safety practices, not a climb to the top.

What Comes Next

For now, Huang appears influential enough to possibly shape policy. He literally has the ear of President Trump, as he demonstrated this week. But the window for industry self-regulation is narrowing fast. Anthropic researchers are already quitting over concerns about self-improving AI. Public trust in tech companies is at historic lows.

The real question isn’t whether regulation will come. It’s whether the AI industry can get ahead of it with meaningful self-imposed safeguards, or whether regulators will step in with blunter instruments that might hamper beneficial innovation alongside risky applications.

Huang is right that you can move fast and maintain safety simultaneously. But leaving it entirely to individual companies betting on their own good judgment, in a race-to-the-bottom competitive environment, might be betting awfully high stakes on awfully optimistic assumptions.

Source: TechCrunch