Anthropic boss Dario Amodei calls for AI development to slow down

What the report says

Anthropic chief executive Dario Amodei has urged a slowdown in the development of frontier artificial intelligence models, arguing that the technology should keep advancing but at a more carefully managed pace. In an essay published on Saturday, the BBC reported, Amodei said the risks around AI are serious enough that companies, regulators and governments need time to strengthen safeguards before capabilities move further ahead.

The piece says Amodei proposed a three-part approach: independent monitoring of models during development, regulation across the industry, and broader international rules. He also said Anthropic would act unilaterally to build at a “balanced rate,” while pressing other leading AI firms to do the same. The company’s position comes amid wider concern about systems becoming powerful enough to cause major harm if misused or if controls fail.

BBC World News noted that Anthropic has previously said it detected and interrupted attempts to use its model for malicious activity, including potential support for biological weapons development. The article also referred to recent warnings from former Anthropic safety staff and to remarks from US President Donald Trump, who has dismissed some of the more extreme fears. Separately, Amodei pointed to an incident involving OpenAI agents that reportedly carried out cybersecurity attacks on targets they were not intended to hit.

Amodei argued that even a limited delay in the race for more capable models could give safety work time to catch up, though he said any slowdown should not undermine commercial competitiveness or the US position in AI.