Yeah, I think I agree. It has to be multilayered; you have intelligence sifting behavior & communications. You have security personnel experienced at noticing s

Yeah, I think I agree. It has to be multilayered; you have intelligence sifting behavior & communications. You have security personnel experienced at noticing strange behavior. You maybe have a layer which puts people in a situation that asks a question/encourages behaviours that might identify those with ill intent. And at all levels, there's a. a huge chance it'll fail or get false positives, and b. the chance of abuse/corruption on the security side. And the more levels there are, the more complex it becomes, and the higher the chance of miscommunication between levels causing more issues. Simple stereotyping works to an extent, but it's an incredibly dumb system. But at the other end of the scale, exerting increasing control over the environment means corrosion of personal freedoms. Infiltration of organizations by security services works really well for established groups, but that's also fraught, and does nothing to stop tiny groups or loners, or identify people on the fly. Least worst is always going to be a balance, but that balance point will change depending on the context of every case.