Predictive policing systems, risk assessment algorithms used in bail and sentencing decisions, and facial recognition tools are all facing intense scrutiny from civil rights organizations. Multiple audits have found that AI systems trained on historical crime data systematically overestimate risk for Black and Hispanic defendants while underestimating it for white defendants.
Several jurisdictions including New York City and San Francisco have passed ordinances restricting or banning certain uses of AI in law enforcement. Federal lawmakers are debating a national algorithmic accountability bill that would require independent audits of all AI systems used in criminal justice contexts. The debate cuts across traditional political lines, with both progressive and libertarian lawmakers expressing concern.
