A critical Trustworthiness Characteristic in the NIST AI RMF, focusing on mitigating the potential for the AI system to produce unjust or discriminatory outcomes against individuals or groups. It requires careful assessment of training data and algorithmic design.

‍