Jensen Huang drew a hard line in the sand at Salesforce’s Dreamforce conference this week. The Nvidia CEO dismissed growing fears that artificial intelligence represents an uncontrollable “alien mind.” Instead, he framed AI as nothing more than complex hardware and software built by humans. Because it is a human creation, Huang argues, it remains under human control. His conclusion? We do not need new regulations. Existing laws and market forces are sufficient to keep dangerous products off the shelves.
“Safety is an engineering problem, not a legal one,” Huang stated. He urged companies to pace their releases based on confidence in product safety rather than regulatory mandates. If a firm isn’t sure its AI is safe, it simply shouldn’t launch. This approach relies on the free market to punish unsafe behavior, a strategy Huang believes allows firms to innovate rapidly without sacrificing security. He called the trade-off between speed and safety a “false choice,” insisting companies can run fast while remaining responsible. It is a comforting narrative for an industry hungry for growth, but it places immense trust in corporate self-policing.
Skepticism is warranted. Huang’s stance aligns perfectly with Nvidia’s financial interests. As the primary supplier of AI chips, any regulatory friction slows down sales. Yet history shows that even well-intentioned tech giants ship flawed code. The 2024 CrowdStrike outage grounded thousands of flights due to a faulty update. Meta recently paid $18 billion to settle claims regarding harms to children on its platforms. AI has already caused tangible damage, from security breaches at Hugging Face to lawsuits linking chatbots to user suicides. Relying solely on product liability laws means waiting for harm to occur before testing legal theories in court.
While Huang pushes for open-weight models as a competitive check on proprietary labs, the window for effective global self-regulation is narrowing. Microsoft CEO Satya Nadella recently noted that safety concerns are universal, urging China to join international standards. But without binding rules, coordination remains voluntary. Huang’s influence is undeniable, evidenced by his direct access to President Trump. If his vision prevails, the AI industry will continue to operate with minimal guardrails, betting that engineering rigor can outpace the unintended consequences of rapid deployment.
