Shieh, E.; Monroe-White, T. Teaching Parrots to See Red: Self-Audits of Generative Language Models Overlook Sociotechnical Harms. AAAI-SS 2025, 6, 333-340.