Police used AI facial recognition to wrongly arrest TN woman for crimes in ND

Police in North Dakota used AI facial recognition to wrongly arrest a Tennessee woman for crimes she didn't commit, exposing a systemic failure even more severe than the typical misidentification case that makes #8's cable tester seem trivial by comparison. The algorithm matched her at just 67 points on a confidence scale, yet officers proceeded with charges that led to a wrongful detention lasting weeks. Over 32 comments on the Hacker News thread highlighted how this case underscores a 90% false-positive rate in certain demographic groups. Unlike the controlled environment of a USB tester reviewed at #8, facial recognition in the wild produces unreliable results that can devastate lives, making this arrest a stark warning about deploying untested AI in law enforcement.