Nvidia CEO Jensen Huang said Tuesday the AI industry doesn’t need new laws or regulations, pushing back directly on a call from Anthropic CEO Dario Amodei just days earlier for the industry to slow down and build in more safety guardrails. The comments put two of AI’s most influential executives on opposite sides of the industry’s biggest open question: whether to regulate itself, or wait for someone else to do it.
What exactly did Huang say?
Speaking at a Salesforce event, Huang said plainly: “We don’t need new laws. We don’t need new regulations.” He argued AI safety is “an engineering problem, not a legal one,” meaning companies should address risk through better testing and design rather than through legislation, and that market forces will naturally discourage companies from releasing products they aren’t confident in. He also dismissed the idea that AI development has to choose between being safe and being fast, calling that framing a “false dichotomy.”
What was he responding to?
Amodei had called for a form of industry-wide coordination on AI safety, which would likely require an antitrust waiver so competing companies could work together on safety standards without running afoul of competition law. Huang rejected that approach directly: “The fact that we need new laws, new antitrust laws, or new regulations, so that these companies could do their fundamental engineering and do it properly before they release products, that is just completely unnecessary.” In Huang’s view, if a company needs a special legal exemption to build its products safely, that’s a sign of a deeper problem with the company’s engineering, not evidence that the law needs to change.
Is Huang’s position really as neutral as “let the market decide”?
Not entirely. Nvidia has been the single biggest financial beneficiary of the AI boom, selling the chips that every major AI lab depends on to train and run its models, so Huang has a direct financial interest in AI development continuing at full speed with as few new legal obstacles as possible. That doesn’t make his engineering argument wrong on its own terms, but it’s worth weighing against Amodei’s opposite financial position: Anthropic is a model developer that would be the one absorbing the cost and risk of moving faster, which gives Amodei’s caution its own kind of self-interest too. Neither executive is arguing from a neutral seat.
Does “no regulation needed” hold up against recent tech failures?
That’s the main counterargument critics are raising. Companies with strong engineering reputations and good intentions have still shipped products that caused real harm, from the 2024 CrowdStrike outage that grounded flights and knocked hospitals offline, to Meta’s settlement over child-safety failures on its platforms. Both cases involved companies that, by Huang’s standard, should have been trusted to self-regulate through engineering discipline and market pressure, and both still produced serious, headline-making failures. Whether AI is different enough from those cases to justify a pure market-forces approach is exactly the debate this exchange has reopened.
Sources: Bloomberg · TechCrunch · Yahoo Finance · Benzinga
