Which statement accurately defines algorithmic bias?

• A. It is a random error in algorithms.
• B. It is a deterministic guarantee of fairness.
• C. It arises from training data quality and design choices.
• D. It cannot occur in text recommendations.

Answer: (C)

Algorithmic bias shows up when a system’s outcomes are systematically unfair due to how it learns from data and how it’s designed to operate. It isn’t random noise; it’s built into the way the model is trained and what objectives, data, and features are chosen. If the training data reflect existing biases or if the design choices place emphasis on certain signals or proxy measures, the system will reproduce or even amplify those biases. For example, a text or content recommender trained on engagement data may prioritize topics popular with a dominant group, reducing exposure for minority voices, unless fairness considerations are built in. This kind of bias can appear anywhere the system ranks, filters, or suggests content, including text recommendations. The other options miss the core issue: bias is not a random error, it is not a guaranteed guarantee of fairness, and bias can indeed occur in text recommendations.