AI is a red herring here (as the article sort of alludes). What should be discussed is what level of privacy invasion should be allowable for law enforcement et

AI is a red herring here (as the article sort of alludes). What should be discussed is what level of privacy invasion should be allowable for law enforcement et al. It seems right now that if an interrogation is non-interactive, it's essentially in bounds. As we get to the point where we can "at scale" harvest information from people that essentially relates to their thoughts and mental state, there needs to be a real discussion of what law enforcement is allowed to do. Any framing of this as "responsible use of AI" or "algorithmic accountability" or whatever is playing right into the hands of the people who want to use this stuff. We shouldn't care how accurate or explainable or unbiased it is, we should care about the limits we place on our governments with respect to monitoring us.