QuestionQ120

AI Risk Management

A hospital is performing an AI impact assessment for a proposed diagnostic-imaging tool. To ensure compliance with regulatory and ethical standards, which risk needs the MOST domain-specific assessment beyond that normally included in a standard IT risk assessment?

Explanation

Systematic bias in diagnostic outputs can cause different demographic groups to receive less accurate diagnoses and can create widespread clinical harm. Assessing this risk requires AI- and healthcare-specific validation of model performance and fairness across relevant patient populations, rather than only conventional infrastructure, security, or data-location controls. NIST’s AI Risk Management Framework identifies fairness with harmful biases managed as a core trustworthy-AI characteristic and calls for fairness and bias evaluation to be documented.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!