In an era where AI automates security questionnaire responses, hidden biases can undermine trust and compliance. This article introduces an ethical bias monitoring engine that works in real time, leverages graph neural networks, explainable AI, and continuous feedback loops to detect, explain, and remediate bias in vendor risk assessments and trust scores.