Nvidia and Anthropic CEOs Split Over AI Safety Pace
TL;DR: At Salesforce's Dreamforce conference, the CEOs of Nvidia and Anthropic offered contrasting views on how to balance the pace of AI development with safety.
Key points
- Nvidia CEO Jensen Huang reportedly called the "go fast vs. slow down" framing a "false choice."
- Huang argued that safety and speed need not be treated as a trade-off.
- Anthropic, by contrast, is reported to have held to its more cautious stance on managing AI risk.
- The two companies represent different layers of the stack — AI infrastructure (chips) versus frontier models — making the split a signal of differing temperatures in the industry's safety debate.
Why it matters
When a leading chip supplier and a frontier model lab publicly send different signals on AI safety and regulatory pace, it can shape upcoming policy discussions and how enterprises approach adoption. It suggests the debate is shifting from a simple "speed vs. regulation" contest toward whether the two can technically advance in parallel.