AI bias refers to systematic errors or unfair outcomes in AI system outputs that result from biased training data, flawed model design, or problematic deployment choices.
AI systems learn from data. If that data reflects historical inequities, underrepresents certain groups, or contains errors, the model may reproduce or amplify those problems. Bias can appear in hiring tools, credit scoring, content moderation, facial recognition, and many other applications.
Bias is not always obvious and may not be visible in aggregate accuracy metrics. Evaluating AI systems for fairness requires examining outcomes across different groups and contexts.