AI could kill all humans in next decade, warn experts: but how seriously should we take them?

Experts and policymakers are increasingly debating the existential risks posed by artificial superintelligence (ASI). Recent warnings from figures like Prof. Stuart Russell and former defense secretary Des Browne highlight concerns that ASI could lead to irreversible catastrophes. Tensions have escalated following the resignation of an Anthropic researcher who claimed companies are racing recklessly toward dangerous capabilities. While some industry insiders estimate a significant probability of human extinction within a decade, others argue these 'Terminator' scenarios distract from immediate, tangible harms like misinformation and job displacement. In the UK and US, calls for international regulation and safety-first treaties are growing, with politicians urging a shift from unconstrained development to rigorous oversight. As labs like Anthropic and OpenAI face pressure to coordinate on safety, the debate remains split between those fearing an existential threat and those dismissing the warnings as alarmist hype.