Artificial Superintelligence, Human Extinction Risk and the Evidence Behind the Growing AI Safety Debate in 2026

A research review of expert warnings, current AI capabilities, loss-of-control scenarios, policy responses, and the gap between speculative catastrophe and documented harm