Explainability-Driven Debugging of Ethical Failures in AI Systems. AI.gov.soc.jrnl. 2025;2(1):19-26. doi:10.5281/