OpenAI's New Safety Policy: Expanded Powers for Internal Safety Committee, First Public Definition of AGI Threshold

OpenAI has announced a major safety policy update: the internal safety committee gains greater authority, the technical threshold for Artificial General Intelligence (AGI) is defined publicly for the first time, and an independent 'Safety Advisory Committee' is established. This article interprets the far-reaching impact of this policy shift on AI industry regulation and OpenAI's strategic adjustments in the AGI race.