UN experts urge science-based approach to AI risk debate

UN experts on artificial intelligence warned Monday against apocalyptic rhetoric about AI risks, urging a more scientific approach to the debate. The warning came during a UN General Assembly event in New York. Google DeepMind executive Joelle Barral said investing in fear is unhelpful, while Nobel laureate Maria Ressa called the debate too polarized between existential doom and irrelevance. The experts responded to recent claims from former AI employees that the technology could kill humanity by the decade's end. Panel co-chair Yoshua Bengio stressed distinguishing known risks from plausible extrapolations, calling for a global rational discussion. The International Scientific Panel on AI, created in 2025, aims to advance scientific understanding and inform UN deliberations.