Legal AI Bias Audit overview

Audits AI systems used in legal workflows for bias and fairness -- evaluating contract review model performance disparities across document types, jurisdictions, and languages, testing e-discovery relevance models for demographic and topic bias in responsive document identification, assessing predictive coding consistency between reviewers and model predictions, reviewing litigation outcome prediction models for protected-class correlation leakage, auditing legal research tool ranking algorithms for citation bias, evaluating explainability and transparency of AI-assisted legal reasoning, and verifying compliance with emerging AI governance regulations (EU AI Act, state-level requirements). Produces bias audit report, fairness metric dashboard, and mitigation recommendations. Excludes model retraining.