Skip to Primary Navigation

Anthropic calls for global ‘pause’ in frontier AI development

A photo of a computer monitor with the webpage for Anthropic's Claude AI platform with the tagline "The AI for problem solvers."
Photo: Michael M. Santiago/Getty Images

As AI self-improving capabilities outpace control frameworks, the tech company wants a time-out to build the tools and framework necessary to govern the technology.

In a rare admission from an industry not much known for self-regulation, tech giant Anthropic warns that fully autonomous, self-improving AI systems “could come sooner than most institutions are prepared for,” and that includes health systems.

A new report from the Anthropic Institute, the company’s research arm, warns that the