Anthropic CEO Dario Amodei has said the time has come to slow down the development of artificial intelligence, according to The Verge. In a lengthy essay, Amodei argued for measures to rein in the speed at which the most advanced AI systems are being built.

Central to his argument is a three-step plan that Amodei described as a way to ‘pace the frontier’ – a phrase The Verge explained simply means to slow the rapid advance of cutting-edge AI models.

As part of the effort, Amodei said Anthropic will give third-party evaluators, including METR, access to its models. The company framed the move as a way to help ensure its ‘adherence to safety practices and commitments,’ according to The Verge.

The commitment to outside evaluation signals an attempt to bring independent oversight to how Anthropic’s AI systems are tested and released. Granting external groups access to models allows them to assess capabilities and potential risks that might not be visible from the outside.

The Verge described Amodei’s essay as ‘winding,’ reflecting a broad set of arguments rather than a single narrow proposal. The core message, however, centered on the idea that the industry should moderate its pace as capabilities grow.

The statements come from the leader of one of the most prominent AI companies, adding weight to a debate over how quickly the technology should progress and what safeguards should accompany it.