Amodei Urges Slower AI Model Development as Huang Defends Safety

Anthropic CEO Dario Amodei has urged a slower pace for advances in AI model capabilities, while Nvidia CEO Jensen Huang says safety and innovation can proceed together.

Amodei proposed ongoing access for independent evaluators, coordination among frontier AI companies in democratic countries on safety standards and development pace, and cooperation between democratic and authoritarian governments. He said Anthropic was committing on its own to provide evaluator access.

Amodei described pacing as allowing time for safety research and outside evaluation to keep up, rather than stopping model training or technical progress. In a September 2026 forecast, he said that within six to 12 months, a misaligned swarm of AI agents might be capable of taking over the internet with a persistent botnet, potentially causing hundreds of billions of dollars in damage.

At a White House event on Sept. 29, Huang said, “There’s no conflict between innovation, technology, and safety.” Amodei said AI carried “very real risks” and that how to address them remained under discussion.

Both executives were among six corporate signers of the White House Accord on Super Intelligence, dated Sept. 29. The voluntary agreement calls for internal controls, an internal oversight team, independent external evaluations and board oversight. It also calls for regular meetings among signers on safety standards and says the controls could later be written into law or regulation.

The accord focuses on company oversight measures and shared standards.