Institutional AI Bias Emerges as New Governance Risk

Risk officers and AI governance leads must audit models trained on proprietary data, since institution-specific outputs can diverge silently from peer systems.