TechUN says AI safeguards can’t wait for certainty

A UN scientific panel says governments should rein in increasingly capable AI agents before their risks are fully understood. Its first thematic brief calls for more resources and stronger international coordination on AI safety and accountability, arguing that scientific uncertainty should not delay safeguards against potentially catastrophic or irreversible harm.

AIUnited NationsAI safety

TThe Verge★★★★☆2026-09-21 10:18Original