A 40-scientist UN panel has determined that AI capabilities are outpacing scientific understanding and government oversight, warning that 'catastrophic harm' cannot be ruled out. The finding marks the first formal AI safety assessment conducted under UN auspices.
July 1, 2026
