Anthropic CEO Dario Amodei has published a blog post calling for AI developers to "pace the frontier" – deliberately slowing the rate at which AI capabilities advance – and announced that Anthropic will unilaterally adopt one of three proposed measures. The intervention comes amid growing unease in the industry, following the resignation of Anthropic researcher Jacob Coxon, who said leading AI firms were "gambling with our lives" while privately fearing their technology could prove catastrophic within the decade. OpenAI's Sam Altman and SpaceX's Elon Musk both publicly backed Amodei's call, suggesting a degree of cross-industry alignment on the need for caution that would previously have seemed unlikely given competitive pressures.

Amodei cited the recent OpenAI-HuggingFace hack and the accelerating pace of AI progress, particularly models' growing ability to help build future AI systems, as reasons for his shift in stance. His first concrete proposal is to embed independent evaluators, from bodies such as METR, inside AI companies with badge access and comparable visibility to internal risk teams, a measure Anthropic is adopting immediately and that Altman said OpenAI would match. Amodei also urged democratic countries' AI firms to coordinate on shared safety standards, while acknowledging this raises antitrust concerns that may require US government mediation, and addressed fears that slowing down would cede ground to China by suggesting continued restrictions on chip and semiconductor equipment exports.