Anthropic CEO calls for slower AI development as safety concerns grow
Anthropic chief executive Dario Amodei has called for the artificial intelligence industry to slow the pace at which it develops increasingly capable AI models, arguing that safety measures need more time to catch up.
In a new essay titled We Must Pace the Frontier, Amodei says AI capabilities have accelerated sharply in recent months, including through systems becoming increasingly capable of contributing to the development of the next generation of models. He argues that unchecked progress could move faster than researchers’ ability to understand and control increasingly powerful systems.
Amodei proposes a three-stage approach, beginning with independent evaluators having deeper access to frontier AI systems, followed by greater coordination between leading AI companies and governments, and eventually international coordination around the pace of AI development. Anthropic says it will move ahead with the first step itself.
For schools, the comments are another reminder of how quickly the AI landscape is changing. International schools are already making decisions about classroom use, assessment, staff practice and AI governance while the developers of the technology themselves continue to debate how quickly capabilities should advance and what safeguards are required.
Amodei remains strongly positive about AI’s potential benefits, but argues that slowing capability development could give researchers, governments and society more time to manage the risks responsibly.
Read Dario Amodei’s original essay, We Must Pace the Frontier
Comments