AI already revolts
From an article in Semafor.com called "Anthropic’s AI resorts to blackmail in simulations" states in a test to the Claude Opus 4 AI system, they created a fictional company with emails saying "it would be replaced, and also that the engineer behind the decision was cheating on his wife". Claude threatened to reveal the affair if the replacement went ahead . Claude Opus 4 is touted as the most intelligent (AI) model to date with advanced reasoning capabilities. Calls for safeguards in AI systems. Too little to late?