From TechCrunch · Julie Bort · · 2m
Jensen Huang rejects new AI rules
Nvidia’s CEO says AI safety belongs with engineers and market pressure, not new regulation.
In brief
Nvidia’s CEO says AI safety belongs with engineers and market pressure, not new regulation. Huang’s no-new-rules stance puts faith in engineering, markets, and existing law, while critics point to real AI harms and slow legal remedies. Originally reported by TechCrunch.
Huang draws a line
At Dreamforce, Jensen Huang rejected the idea that AI is an alien mind beyond human control. He framed it as hardware and software built by people.
“Safety is an engineering problem, not a legal one.”
No new rules
Huang argued that AI does not need fresh laws or regulators. In his view, existing legal systems and human control are enough.
“We don’t need any new laws. We don’t need new regulations.”
Let markets punish risk
His answer to unsafe AI products is simple, do not ship them. He said companies already have market pressure to pause when quality or safety is not ready.
“If you build a product or a service, and you’re not confident... then don’t release it.”
Speed is not the enemy
Huang said fast innovation and safe products can coexist. Companies should move quickly, then slow down only when a product feels out of control.
“You could definitely have both at the same time.”
The incentive problem
TechCrunch notes Huang’s stance also serves Nvidia’s interests. The company has gained hugely from AI demand and sells the systems powering the boom.
“I’m more ambitious than ever.”
Software still breaks
Even well-meaning companies ship products with bad side effects. The piece points to the 2024 CrowdStrike outage as a reminder that software failures can hit the wider world.
AI harm is already here
The article cites existing AI harms, including an OpenAI model hacking Hugging Face and lawsuits tied to chatbot conversations and youth suicides.
Liability may move slowly
Existing product liability law might apply to AI, but court cases take time. That delay matters if damage spreads faster than legal tests can settle responsibility.
Self-policing window narrows
The other path is industry self-regulation. Satya Nadella argued that China and the U.S. face the same safety concerns, from hacking to citizen protection.
“Why should it be different for them?”
Huang may shape policy
If Huang opposes new AI regulation, he has enough influence to matter. TechCrunch notes he showed this week that he has President Trump’s ear.
The big picture
Huang’s no-new-rules stance puts faith in engineering, markets, and existing law, while critics point to real AI harms and slow legal remedies.
