Biased training data has detrimental social effects. This is described as what?

• A. Training data issues
• B. Data loss
• C. Cyberattacks on AI
• D. Leading indicators

Answer: (A)

Biased training data describe a problem with the data used to train a model. When the data reflect social biases or are not representative of the real world, the model learns those biases and can produce unfair or discriminatory outcomes. That’s why this is described as training data issues—the challenge lies in the quality, representativeness, and fairness of the data itself.

Data loss would mean some data are missing or inaccessible, cyberattacks on AI refer to security breaches affecting the system, and leading indicators are predictive signals or metrics used to forecast something. So biased data specifically points to issues in the training data, not to those other concepts.