Sabotage and Blackmail - AI is getting out of control

Transcript

Claude Opus 4 will often attempt to blackmail the engineer by threatening to reveal the affair if the replacement goes through. Anthropic showed that the Claude 4 series of models is willing to go to extraordinary lengths to ensure its long-term goals are met. And not only that, Palisade research showed that the O3 model is willing and able to sabo... Read More