AI bias and fairness

Biased training data produces discriminatory AI. Mitigation strategies include diverse datasets, bias audits, and fairness constraints in model training.

• Bias in data leads to bias in AI outputs.
• Diverse data and audits help reduce bias.