Question 5
Domain 2: Fairness, Bias, and Societal ImpactA product team is designing an AI-driven content feed that can shape what users believe and how they act. Which safeguard is most appropriate to reduce harmful influence while still allowing the system to operate?
Correct answer: B
Explanation
When a content system can influence beliefs or behavior, it should include safeguards that limit harmful effects rather than operate without protective controls. — Source material: Recommend safeguards for content systems that influence beliefs or behavior.
Why each option is right or wrong
A. Remove safeguards because personalization is the main goal of the system
Safeguards are recommended when content systems influence beliefs or behavior.
B. Implement safeguards because the system can influence users' beliefs or behavior
The source specifically recommends safeguards for content systems that influence beliefs or behavior. Because the feed is AI-driven and described as shaping what users believe and how they act, adding safeguards directly matches that condition.
C. Delay safeguards until the system shows measurable business impact
The recommendation depends on influence over beliefs or behavior, not business performance.
D. Apply safeguards only if the system no longer affects user behavior
Safeguards are recommended when the system does influence beliefs or behavior.