UN experts urge nuanced AI risk debate at UN panel event

Experts on a United Nations artificial intelligence committee are cautioning against alarmist, "apocalyptic" rhetoric about AI risks, urging a focus on scientific understanding rather than fear-based speculation. Speaking at a New York event on September 21, 2026, alongside the UN General Assembly, Google DeepMind executive Joelle Barral stressed the importance of distinguishing known facts from unknowns. The warnings follow recent claims by former AI industry workers that the technology could "kill us all by the end of the decade." The International Scientific Panel on AI, created in 2025 with members selected in February 2026, aims to advance scientific understanding of AI to inform UN deliberations. Co-chair Yoshua Bengio emphasized differentiating between established knowledge and plausible extrapolations, while Nobel laureate Maria Ressa called for changing the polarized discourse around AI risk.