Speaking at Salesforce’s Dreamforce conference on Tuesday, Nvidia founder and CEO Jensen Huang made his position on the dangers posed by AI clear. To him, AI is not a new form of “alien mind,” as at least one OpenAI safety researcher has described it. It’s just hardware and software made by humans, he says. That means, in his view, that it can be controlled by humans and existing laws.
“Safety is a technical issue, not a legal issue,” he says. “At the end of the day, we’re building software. At the end of the day, we’re building computing systems. It’s a complex computing system, but ultimately it’s a computing system.”
Therefore, he argues, there is no need for new laws or regulations to manage it. In fact, Huang sees no need for new laws at all. He argues that the free market is enough to pressure companies not to release unsafe products.
“Like you and me and all the companies here, if you don’t have confidence in the safety of your product, if you build a product or a service and you don’t have confidence in its functionality or its capabilities or its safety, don’t release it. So it makes perfect sense,” he said.
“You just have to go at your own pace until you’re confident you’re releasing something that the market appreciates. The market forces are already there. We don’t need new laws. We don’t need new regulations. We just need companies to decide when to do it and as quickly as possible. I think it’s a false choice. So do it as soon as possible. But if you feel like the company is out of control at a certain point, then stop and see if you get it right.”
In a way, this is a comforting thought. If there’s anyone in the world who knows AI, it’s the founder of Nvidia. He was building the hardware brain for AI long before ChatGPT existed, and now runs a company that also creates open source models, agents, harnesses, and sandboxes.
Again, to be cynical, his perspective is not surprising for someone who has been able to make a good living from the AI boom. Why would he want regulations introduced that would add a layer of hurdles that could slow down Nvidia’s efforts to sell ever more AI systems and software? He also said in an interview: “I am more ambitious than ever. Because of our ambition and the productivity of our products through AI, there are no limits for us. There are no limits for us. There are no limits for every industry and every country.”
Unfortunately, even well-intentioned companies and software end up shipping defective products that have unintended consequences. Remember the deadly CrowdStrike blue screen crash in 2024 that grounded thousands of flights and caused further havoc for businesses, some of which were even accused of acting intentionally with less-than-beneficent intentions? Meta just paid $18 billion to settle a lawsuit over social media harm to children.
And whatever the intentions or safety tests, AI is already causing harm, from the OpenAI model that hacked Hugging Face to a lawsuit against an AI institute over the suicide of a young man who had an extended conversation with a chatbot.
A “let alone” strategy that allows these companies to release the products they deem necessary may be an unwise approach to AI safety as far as society is concerned. Although Huang is correct in pointing out that existing product liability laws may cover AI, unless it somehow kills us all before enough cases get through the courts to test that theory.
He did not mention another route that appears to be taking shape: industry self-regulation. Huang’s approach championed the open-weight model and its use by companies as a competitive counterweight to their own AI labs.
But for now, there is a short window for the industry to adopt self-regulation and encourage intelligent participation from AI labs around the world, including China. As Microsoft CEO Satya Nadella said at Monday’s All-In Summit, “If the United States is concerned about security, then why should China be deeply concerned about similar security concerns? Why should it be any different for China? It’s not that they don’t have the same hacking problems. Just as we want our citizens to benefit from AI, it’s not that China doesn’t want to make sure its citizens benefit from AI.”
But for now, if Hwang opposes new AI regulations, he may have enough influence to get his way. He proved this week that he literally has President Trump’s ear.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
