Jump to content

AI Law Wiki News for September 12, 2026

From AI Law Wiki

AI-generated text. This page was generated using artificial intelligence.

September 12, 2026Anthropic’s Dario Amodei proposes slower frontier-AI development and commits to embedded outside safety evaluators.[1] Hugging Face’s Clément Delangue announces an Open Alignment Initiative and seeks a role in the evaluator program.[2] OpenAI’s Sam Altman also commits to independent evaluators with access comparable to employees.[3]

Amodei proposes pacing AI development and embedded evaluators

Anthropic chief executive Dario Amodei published a proposal to slow capability advances while improving safety, according to CNN’s September 12 report.[1]

His three-part framework calls for embedded third-party evaluators, coordination among companies in democracies, and international agreements. He says Anthropic is committing to the first step; the others require wider cooperation. He distinguishes pacing from stopping model training.[4]

The proposed evaluators would receive ongoing access comparable to employees and could publish findings, subject to limited confidentiality and security redactions. Amodei also advocates regulation and government assistance with legally sensitive industry coordination; these are proposals, not newly enacted requirements.[4]

On September 12, 2026, OpenAI chief executive Sam Altman endorsed Amodei’s pacing call and said OpenAI would also give independent evaluators access comparable to employees, with further details to follow. This was an announced commitment, not confirmation that the arrangement was already operating.[3][1]

On September 12, Clément Delangue announced the Open Alignment Initiative, led by Thomas Wolf at Hugging Face, and requested participation in the embedded-evaluator program. He argued for broader participation and transparency in alignment work; the announcement does not confirm acceptance into the program.[2]

On September 12, 2026, former UK prime minister Rishi Sunak, who identified himself as an Anthropic senior adviser, supported Amodei's pacing call. He argued that voluntary pre-release testing was no longer sufficient and urged government-backed independent verification, model security and US–China dialogue. These were policy recommendations, not newly enacted requirements.[5]

See United States policy on catastrophic AI risk for the broader policy debate.

References