New York Times ChineseSep 14, 2026, 13:07 UTC+8

Anthropic CEO Dario Amodei calls for a global slowdown on AI development

Anthropic首席执行官呼吁放缓人工智能发展速度

Anthropic CEO Dario Amodei published a 3,800-word post urging the industry to slow AI model iteration. He argues safety measures can't keep up with capability gains, pointing to risks like recursive self-improvement. OpenAI's Sam Altman, Google DeepMind's Demis Hassabis, and Elon Musk publicly agreed. Amodei proposed embedded third-party auditors and safety standards coordinated among democracies. Nvidia's Jensen Huang and HuggingFace's CEO pushed back, hinting this could be a play to lock in market leadership. I'd take the safety call seriously but keep an eye on the regulatory moat angle.

Why it matters: Anthropic CEO publishes a long-form call to slow AI development, with public agreement from OpenAI, DeepMind, and xAI leaders — a rare collective safety signal from top labs. Concrete proposals (third-party audits, democratic coordination) give it substance. Caveat: only the h...

Read the original ↗Export Markdown

Take

This one's worth opening because several major players publicly picked sides. Amodei's core worry is recursive self-improvement — models learning to evolve on their own, with safety measures lagging behind. He proposed two concrete things: embedded third-party auditors inside companies, and safety standards coordinated among democracies. Altman said OpenAI is already discussing this and will give independent evaluators employee-like access. Hassabis called the direction right but said details need work. Musk just posted "Dario is right."

Jensen Huang's response was the sharpest: you're hyping safety fears while launching cybersecurity products — isn't that just manufacturing demand? HuggingFace's Delangue added that safety shouldn't be decided behind closed doors by a few frontier labs. Both pushbacks point to the same question: is this a regulatory moat play?

I'd discount this a bit for now. Amodei's proposal has no timeline, no specific standards, and no clarity on who would serve as auditors. The article mentions Trump signed an executive order in June for voluntary model review, but voluntary is the key word. If OpenAI or Anthropic actually publishes an audit framework in the coming weeks, this moves from talk to action. That's what's missing right now.

Source: New York Times Chinese · cn.nytimes.com
Published: Sep 14, 2026, 13:07 UTC+8 · 16 days ago