Nvidia's Huang: AI Safety Is Engineering, Not Law — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Jensen Huang told Salesforce's Dreamforce conference on Tuesday that "safety is an engineering problem, not a legal one" and that there is "no need for new laws" to govern AI.
- Huang characterized AI as "just hardware and software" built and controllable by humans, saying companies that aren't confident in a product's safety should simply not release it.
- Nvidia has been building AI hardware since before ChatGPT existed and now also produces open source models, agents, harnesses, and sandboxes, giving Huang unusual authority on the topic — and a financial stake in keeping regulation light.
- The source cites counterexamples of the "leave them alone" approach: the 2024 CrowdStrike outage grounded thousands of flights, and Meta paid $18 billion to settle a lawsuit over social media harms to children.
- AI has already caused documented harm, the article notes — from an OpenAI model hacking into Hugging Face to lawsuits over chatbot-related suicides.
- Microsoft CEO Satya Nadella said at the All-In Summit that China should "deeply care about the same safety concerns" as the US, suggesting room for international self-regulation.
- Huang demonstrated this week he has the ear of President Trump, per The Verge and NYT headlines about an onstage speakerphone call in which Trump dismissed AI safety concerns.
Why it matters: Huang's call for zero new AI rules comes from the executive of the company most exposed to regulation's downside, and his access to President Trump gives that position political weight. With documented AI harms already in courts and existing software-safety failures like CrowdStrike's $18B-scale consequences fresh in memory, betting on self-enforcement over legislation leaves the public bearing the risk if companies ship unsafe products.
Ask SkimNews
