Citizens are skeptical about letting AI make welfare decisions, especially those who actually depend on social services, according to new research that reveals a troubling gap between policymakers' ...