Raw LLM Responses

I do worry that this particular case will be used later as a supporting evidence for "police _should_ be allowed to use face recognition to identify suspects" argument. Given how swiftly all the arguments against such a thing get abandoned when tool is provided to the public. "See, you honor, public _loves_ this kind of service and potentially wrong matches never seem to be a worry. They'd rely on a random web site claiming to get the officer's badge number, get wrong data back and it'll be taken at a face value to harass the wrong guy, while when police uses face ID there's an extra verification step" type of thing.