Are we heading towards police states? Not if we don't want to, probably. Problem is that policemen have a professional education at best, don't expect PhDs there. They are not statisticians.
All statisticians know that digital data, given large numbers, will only generate false positives, and all real positives are statistically insignificant. The only thing one can do with digital data is falsify real-world positives; how hard is it to get that notion across?
