Question 16
Domain 4 — AI Governance, Risk, Ethics, and TrustworthinessEthical Perspectives in AI Course: A company is developing an AI system for hiring decisions. During testing, it is discovered that the AI favors certain demographic groups over others due to biased training data. What should the company do to address this ethical issue?
Correct answer: B
Explanation
Retraining the AI with more diverse and representative data addresses the bias at its source, since biased training data can produce discriminatory outcomes. Fair hiring systems should avoid favoring groups based on protected characteristics, so improving the dataset helps reduce unequal treatment and supports more ethical decision-making.
Why each option is right or wrong
A. Ignore the bias since it reflects real-world data.
B. Retrain the AI with more diverse and representative data.
Under anti-discrimination hiring principles, an AI tool that produces disparate treatment or disparate impact from biased inputs must be corrected at the source rather than left in deployment. Retraining with a broader, more representative dataset is the appropriate remedial step because it targets the training bias that is causing the model to favor certain protected groups, which is especially problematic in employment decisions where fairness and equal opportunity are required.
C. Limit the AI's use to non-critical hiring decisions
D. Replace the AI system with human decision-makers