Nvidia: Huang rejects new AI legislation

Jensen Huang, founder and CEO of Nvidia, spoke on September 16, 2026, at Dreamforce about the current and future regulations surrounding AI. Huang argued that additional legislation for artificial intelligence is unnecessary, and that companies themselves are responsible for ensuring safety by using technology within existing legal frameworks. He sees AI as a combination of hardware and software, built by humans, and believes that market mechanisms already compel companies not to release unsafe products.

According to Huang, the issue of safety in AI is not a legal matter, but a technical one. He emphasized that if a company has doubts about the functionality or safety of its product, it should not be released. He highlights this as an obvious choice for companies in the technology sector. He also looked at the speed at which companies operate and stated that activity and safety do not have to be opposites. If a company senses that a product may be unsafe, it must immediately stop and evaluate the matter.

Huang referenced past incidents where software issues, not necessarily AI, had significant consequences. Examples include chaos in the aviation industry and damage claims against large companies like Meta, which paid $18 billion for a settlement related to social media. He points to the reality that AI has also led to legal disputes, but emphasizes that the responsibility for safety primarily lies with the companies themselves.

A consequential discussion that Huang raised was that of self-regulation in the sector. He proposed that companies adopt so-called open-weight models as an alternative to the closed models of traditional AI labs. This would allow companies to compete and innovate without the need for additional regulations. Huang's positions raise questions about the feasibility of collaborative safety standards and the role of governments in benefiting from AI. Furthermore, Huang also made it clear that he continues to pursue his ambitious plans for Nvidia, which may influence the political debate surrounding AI regulation.

Read the full article from AI Insider.