Nvidia founder and CEO Jensen Huang made his views on AI safety and regulation clear while speaking at Salesforce’s Dreamforce conference on Tuesday.

 

Huang rejected the idea that AI represents a new kind of “alien mind,” arguing that it is ultimately hardware and software created by humans. Because of that, he believes humans can control AI using engineering practices and existing laws.

 

“Safety is an engineering problem, not a legal one,” Huang said. He described AI as a sophisticated computing system rather than something fundamentally different from other technologies.

 

Based on that view, Huang argued that new AI-specific laws and regulations are unnecessary. Instead, he believes market forces can push companies to prioritize safety and avoid releasing products they are not confident in.

 

“If we’re not confident about the safety of the products… then don’t release it,” he said, adding that companies should slow down when necessary to make sure their products are safe.

 

Huang also rejected the idea that innovation and safety are mutually exclusive. In his view, companies can move quickly while still pausing when a product becomes difficult to control or raises safety concerns.

 

His position is understandable given Nvidia’s central role in the AI industry. The company has become one of the biggest beneficiaries of the rapid growth in AI computing and is also expanding into AI models, agents, software, and other technologies.

 

However, critics of a purely voluntary approach point out that even well-intentioned companies can release products with unexpected consequences. The 2024 CrowdStrike outage, for example, demonstrated how software failures can create widespread disruption. There are also cases involving companies accused of putting commercial interests ahead of safety.

 

AI systems have already generated concerns over cybersecurity, misinformation, privacy, and other potential harms. These examples raise questions about whether relying primarily on companies to decide when their own products are safe is sufficient.

 

Huang’s position also leaves open an important question: whether existing product-liability and other laws are enough to address risks created by increasingly capable AI systems. The effectiveness of those laws will depend partly on how courts and regulators apply them to emerging AI technologies.

 

Another possible approach is industry self-regulation, where AI companies establish common safety standards and practices without waiting for governments to impose new rules. Such an approach could require cooperation among major AI companies and across countries.

 

Microsoft CEO Satya Nadella recently emphasized that AI safety concerns are not limited to one country, arguing that companies and governments worldwide face many of the same risks.

 

For now, Huang is clearly advocating a model centered on engineering, responsible product development, and market incentives rather than additional AI-specific regulation. His influence in the technology industry—and his growing access to policymakers—could make his position an important part of the ongoing debate over how AI should be governed.