Anthropic, a leading artificial intelligence lab, is calling for a coordinated and verifiable pause in AI development, warning that rapid advances could soon allow AI systems to improve themselves faster than society can manage the risks.
The lab's ability to complete tasks on its own has been doubling roughly every four months, and it's headed for 'recursive self-improvement,' the point at which the technology can improve without human intervention.
Anthropic co-founders Jack Clark and Marina Favaro wrote in a blog post that a pause would allow society to 'deal with its immense implications' and that regulation has been slow, especially in the U.S. where most leading AI labs are based.
The lab has long positioned itself as a safety-focused AI lab and has refused to let the U.S. military use its models for domestic surveillance and fully autonomous weapons, prompting backlash from the government.
Anthropic's Thursday post cautioned that unilateral or poorly coordinated slowdowns could backfire if less cautious actors continue advancing, potentially reducing overall safety.
The lab's research arm, Anthropic Institute, plans to study systems needed to support a slowdown and will convene policymakers, researchers, civil society groups, and rival AI firms to discuss managing risks such as recursive self-improvement.
