AI Weekly - OpenAI Admits Its Safety Net Is Fraying

Something shifted this week. The frontier labs stopped defending their controls and started describing how they are failing - and Parliament began drafting the instrument it would reach for. OpenAI...