Anthropic chief executive calls for deliberate pacing of AI capabilities amid documented misuse risks.
Anthropic CEO Dario Amodei on Saturday, September 12, 2026, published an essay titled "We Must Pace the Frontier" on his personal website, darioamodei.com, calling on AI companies to slow the rate at which they advance model capabilities, according to Reuters and The Guardian. Amodei amplified the essay in a post on X the same day, describing a three-part plan he said was designed to address mounting fears over AI misuse.
The Three-Part Framework
View this post on Instagram
Amodei's proposal centers on three steps: deep third-party evaluation inside frontier labs, coordination among leading AI companies on shared safety standards, and international cooperation on governance mechanisms, according to Reuters and Amodei's own essay. Anthropic said it would unilaterally commit to the first step by granting third-party evaluators "permanent, employee-level access" to its systems, allowing them to verify safety measures, report incidents and assess model alignment during training, Amodei wrote on X. "We must slow the pace at which we improve the capabilities of AI models. Progress will still seem fast, and we must make wise use of the time we gain," Amodei stated, as quoted by Reuters and The Guardian. He clarified he was not calling for a full halt to AI training or research.
Google Removes Paywall From Dreambeans AI Feed Across United States
Misuse Findings and Earlier Warnings
Reuters reported that Amodei's essay followed Anthropic's release of a threat intelligence report detailing how various actors used its Claude models for activities including weapons-related research, cyber operations, surveillance and fraud. The September call builds on an earlier warning from June 2026, when Anthropic co-founder Jack Clark and Marina Favaro, head of the Anthropic Institute, urged major AI labs to consider a coordinated and verifiable pause in development, cautioning that advanced systems could eventually "fully build their own successors," AP News reported. Anthropic said at the time that a coordinated pause would be beneficial but stressed such capability was "not inevitable."
As of publication, no industry-wide agreement on Amodei's framework has been confirmed. Reuters noted the essay arrives amid broader regulatory scrutiny of AI, including a Senate probe into an OpenAI-linked incident and federal agencies reassessing AI vendor relationships, underscoring wider institutional concern over the pace of AI development.
