Every AI disaster scenario has already happened somewhere without AI. A human caught each one
What happened
Chatbots haven't made bioweapons or water attacks easier to pull off. Their real danger is a false report that misleads someone with minutes to act
Summary assembled by rule from the sources below