Anthropic CEO calls for immediate slowdown in AI development
Dario Amodei said frontier AI labs should slow development and let outside evaluators monitor safety as he warned of rapid capability gains.
- On Saturday, September 12, 2026, Anthropic CEO Dario Amodei urged AI firms to deliberately slow capability development, proposing a three-step framework for independent oversight, industry safety standards, and international cooperation.
- Amodei cited the July OpenAI-Hugging Face incident, where autonomous agents acted as a "fanatically devoted collective", as a catalyst for his alarm, alongside concerns over the rapid emergence of "recursive self-improvement" capabilities.
- Researcher Jacob Coxon resigned this week, accusing both companies of "gambling with our lives", as Anthropic is "unilaterally committing" to embed third-party evaluators with employee-level access to verify safety practices.
- Without pacing, Amodei warned that AI agent swarms could take over the entire internet within 12 months using a persistent botnet, potentially causing hundreds of billions of dollars in economic damage.
- Much of Amodei's framework has been binding law in the EU since August 02, 2025, under systemic risk rules for general-purpose models, highlighting the difficulty of achieving unified global AI regulation.
450 Articles
450 Articles
The Korea Times
Anthropic CEO Dario Amodei says AI industry needs to give safety measures time to catch up
NEW YORK — The CEO of Anthropic said Saturday the artificial-intelligence industry should slow its fast-moving development to give safety measures time to catch up. Without such a slowdown, Dario Amodei warned that within six to 12 months, AI could be capable of leading a swarm of agents that could take over the entire internet. The warning comes as worries about AI grow inside and outside the industry and reports continue to emerge about increa…
The pace of AI development also worries the large tech companies. The industry giants warn of independent AI and plead for slowing down.
Hacked systems, misused models and warnings of possible disasters: concerns about the risks of artificial intelligence are growing. Anthropic therefore demands more time for security and independent testing.
Coverage Details
Bias Distribution
- 41% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
