Jensen Huang Tells Dreamforce AI Safety Is an Engineering Problem, Not a Legal One

During the Dreamforce conference hosted by Salesforce on Tuesday, Jensen Huang, the CEO of Nvidia, asserted that AI consists of controllable hardware and software developed by humans. He remarked, "Safety is an engineering problem, not a legal one," and argued against the need for new regulations, suggesting that market forces will deter unsafe products. Huang highlighted the importance of rapid innovation while maintaining safety standards, noting that AI-driven productivity offers significant opportunities for Nvidia and various sectors. He referenced incidents of AI failures, such as the CrowdStrike outage in 2024 and lawsuits against OpenAI. Additionally, Huang supported open-weight models instead of proprietary alternatives. Microsoft’s CEO, Satya Nadella, noted that China shares the U.S. concerns regarding safety, while Huang's influence could potentially prevent new AI regulations.