I think this is why I don't think slander would work. Clearview AI didn't force the cops to arrest him, or say with 100% accuracy it was him. CVAI could have

I think this is why I don't think slander would work. Clearview AI didn't force the cops to arrest him, or say with 100% accuracy it was him. CVAI could have an accuracy of 99.9%, and the defense would be "Well see you were that 0.1% incorrect!" It's the police that need to do the investigation to make sure the guy arrested is the same person on the video. But then they could say too, "Well CVAI said it was a near certainty -- we just took their word for it! It would have wasted the taxpayers money to do a more thorough investigation -- afterall there's always some implicit uncertainty in standard policework!" I think a more interesting question is something like: Does Clearview AI fudge their accuracy numbers? Would the true 80% likelihood of you being the perp be more or less slanderous than a 99% likelihood?