hey anthery, good question. tbh, it really depends on what kind of 'specific suspicious behaviors' you're thinking of.
for pretty clear-cut stuff like someone loitering in a restricted zone for too long, or a car stopping at an odd hour, yeah, a lot of the newer AI cameras are pretty good at flagging those. some systems are also getting decent at things like detecting falls (useful for elderly monitoring) or even recognizing a package being picked up from a porch. the accuracy there is definitely improving.
but if you're talking about more subtle stuff, like trying to predict intent or recognizing really complex human behavior patterns from a distance, nah, not quite there yet imo. you still get a fair amount of false positives too – like a delivery driver looking 'suspicious' when they're just doing their job, or an animal setting off an alert.
it's a powerful tool for basic automated security and monitoring, but it's not a complete replacement for human observation or analysis, especially for high-stakes situations. the tech is definitely evolving fast though!
