Could AI Trigger Doomsday? Experts Map Possible Pathways

AI safety researchers warn that powerful, possibly self-improving systems could cause catastrophic harm—from rogue agents hijacking critical infrastructure to AI-driven coups in government—through scenarios like malicious use, race dynamics, and organizational failures. The debate spans bioweapons design, nuclear command risks, and rapid deployment without adequate safeguards, with some leaders advocating safety-focused pacing while others fear regulatory lag.