When AI Facial Recognition Gets It Wrong, Who Pays the Price?

Facial recognition technology fails dramatically across demographic groups. At least eight Americans have been wrongfully arrested due to facial recognition misidentifications. NIST studies of 189 algorithms found most systems falsely identify Black and Asian faces 10 to 100 times more often than white faces. Despite industry claims of 99%+ accuracy, real-world deployment reveals systematic failures that have turned everyday technology into a source of civil rights violations and wrongful arrests.