英国议会 AI 安全听证：对齐、事件报告与实验室问责

The UK Parliament committee holds a full hearing on AI safety and governance. The publisher lists three witness groups: Daniel Kokotajlo, Aengus Lynch, Steven Adler, and Geoffrey Irving; policy leaders from Anthropic, OpenAI, Google DeepMind, and Meta; and officials from the AI Security Institute and UK government.

The hearing addresses model safety, risk controls, and public oversight. Researcher testimony, corporate responses, and government statements should be distinguished. The recording was published on October 13 and is added retrospectively. Only automatic captions are available in the private collection; an official transcript and individual quotations have not been verified.