AAIR: Advanced in AI Risk

By Isaca · Question Mode

QuestionQ38

VerifiedAI Life Cycle Risk Management

An organization deploys an AI credit-scoring model trained on historical financial data that underrepresents certain demographic groups. Which option is the risk practitioner’s BEST recommendation for mitigating this risk?

Explanation

Representative and inclusive training data reduces the risk that a credit-scoring model systematically disadvantages demographic groups that were underrepresented in its historical data. Defining explicit inclusivity goals and expanding the data sources directly addresses the underlying representation bias.

Community Discussion

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