Illustration by Tag Hartman-Simkins / Futurism. Source: Getty Images These days, AI surveillance systems are everywhere. They’re watching our roads, tracking our purchases at the grocery store, and even monitoring school bathrooms. Yet those systems come with a harmful catch — they’re a little too good at flagging bad behavior, leading to a staggering number of arrests and involuntary detentions. New reporting by the Milwaukee Independent found that schools deploying surveillance systems to watch for online chatter referencing violence have logged an astonishing number of positive hits. Many of them are obviously not actionable threats, though they can lead to involuntary detention all the same. In one school district in Lawrence, Kansas, for example, an online safety monitoring tool called Gaggle flagged over 1,200 online incidents over a 10-month period. Yet two thirds of those alerts were nonissues, the Independent found. While many false positives can be waved away, schools in states with zero-tolerance reporting laws like Tennessee or Florida are required to inform police officials about cases where students hint at violence in any …