Policy Design CRAGE Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   In AI governance policy design, which principle requires that the operation and decision-making of AI systems be understandable to relevant stakeholders?

   • AFairness
   • BTransparency
   • CAccountability
   • DPrivacy
   Reveal Solution Discussion Select an answer first
2. 22application · medium

   A university is developing an AI system to screen student applications. The policy team wants to engage stakeholders to ensure the policy is fair and accepted by the community. Which stakeholder engagement method is most likely to surface concerns about bias?

   • AHold focus groups with current students, alumni, and admissions staff, and include a structured discussion on fairness and bias.
   • BSend a survey to all applicants asking if they have any concerns about the AI system.
   • CConsult only with the admissions office, since they are the primary users of the system.
   • DReview academic literature on bias in admissions AI and incorporate findings into the policy.
   Reveal Solution Discussion Select an answer first
4. 23application · medium

   A city government is drafting an AI policy for a predictive policing system. Community advocates have raised concerns about bias and civil liberties, while the police department wants operational flexibility. The policy team must ensure the policy reflects diverse perspectives and gains buy-in. Which engagement method is most effective?

   • APublish a draft policy on the city website and invite written comments within a 30-day period, with no other outreach.
   • BConduct a series of public workshops and structured interviews with community groups, civil rights organizations, and police officers, then incorporate their feedback into a draft policy for public comment.
   • CForm an internal committee of city IT staff and legal advisors to finalize the policy without external input.
   • DRely on a third-party consultant to write the policy based on industry best practices, with no direct stakeholder consultation.
   Reveal Solution Discussion Select an answer first
5. 24application · medium

   A government agency is designing an AI policy for a welfare eligibility system. The system will automate decisions about benefits. The policy must incorporate ethical considerations such as human dignity and the right to appeal. Which policy element is most essential?

   • AMaximize the system's efficiency by reducing the time to process applications, with no additional review steps.
   • BEnsure that every automated decision can be reviewed by a human caseworker upon request, and that the policy includes a clear appeals process.
   • CPublish the algorithm's decision criteria to the public to ensure transparency, but do not provide individual explanations.
   • DUse a highly accurate AI model and assume that errors are negligible, so no appeals process is needed.
   Reveal Solution Discussion Select an answer first
6. 25expert · hard

   A retail company has implemented an AI governance policy that requires regular audits of its pricing algorithms. The audits have revealed a pattern of price discrimination against certain customer groups. The company's leadership wants to fix the issue quickly, but the audit process is slow and the next audit is not scheduled for six months. Which enforcement strategy best addresses the urgent need for remediation while maintaining the integrity of the audit process?

   • ADisable the pricing algorithms entirely until the next audit is completed.
   • BWait for the next scheduled audit to avoid disrupting the audit schedule.
   • CTrigger an immediate targeted audit of the affected algorithms and require a remediation plan within 30 days.
   • DConduct an informal review by the data science team without involving the audit function.
   Reveal Solution Discussion Select an answer first