A risk assessment team discovers that an AI model used for employee performance reviews was trained using data from only one regional office, which is predominantly male. What is the PRIMARY risk this introduces?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — training on data from a demographic monoculture means the model may not generalize to female employees and could produce discriminatory performance predictions. That's sampling bias, and it's the primary risk here.
Full Explanation
The correct answer is sampling bias that may produce discriminatory performance predictions for female employees. Training on data from a demographic monoculture means the model's learned patterns may not generalize to female employees, potentially producing systematically lower or inappropriate performance scores. This constitutes sampling bias with a direct discrimination risk. The overfitting risk due to a small training sample size is a performance concern, but it is secondary to the bias concern. The concept drift risk because performance standards differ across regions applies to temporal shifts, not regional representativeness issues. The label noise risk from inconsistent manager ratings across locations may coexist, but it is not the primary risk identified.