AI Hallucination Risk Research Archive