Which of the following best describes how bias and fairness concerns are addressed in AI advertising, and what mitigation steps are recommended?

• A. Biased models can discriminate or misrepresent groups; mitigate with diverse training data, fairness metrics, bias audits, and human-in-the-loop governance.
• B. Bias does not affect advertising outcomes; no mitigation needed.
• C. Fairness can be ensured by removing all demographic features from data; this fully eliminates bias.
• D. Fairness metrics are unnecessary if accuracy is high.

Answer: (A)