In Torts, I’m getting many questions about the potential liability of AI for giving negligent or reckless advice. My initial response to these questions is to ask “How would you analyze the case if the advice were given by a person?”
This is a useful starting point because it pierces the AI distortion field. Sometimes the answer is that AI and human advice should be judged alike. Sometimes they seem different, though we find it hard to articulate why. As people make that attempt, though, the contours of AI tort liability will emerge.