Policy Design CRAGE Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A social media company is designing an AI policy for content moderation. The policy must respect freedom of expression while also protecting users from harmful content. The company operates globally, including in jurisdictions with strict content laws. Which policy approach best balances these ethical and legal considerations?

   • AUse an automated AI system to remove content flagged by the algorithm without human review, to ensure speed and consistency.
   • BAdopt the strictest content standards from any jurisdiction and apply them uniformly worldwide to ensure compliance everywhere.
   • CCreate separate policies for each country based solely on local laws, with no common ethical framework.
   • DDevelop a policy that defines clear categories of harmful content, includes human review for borderline cases, and aligns with regional legal requirements while maintaining a global baseline of human rights standards.
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   A retail company uses an AI system to personalize marketing emails. The system processes customer data, and the company operates in the EU. The policy team is aligning the AI policy with GDPR. Which policy element is most directly required by GDPR?

   • AConduct a bias audit of the AI system every year.
   • BProvide customers with a clear privacy notice explaining how their data is used for personalization and how they can opt out.
   • CEnsure the AI system is deployed on-premises to avoid data transfers outside the EU.
   • DPublish the AI algorithm's source code to ensure transparency.
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   What is a primary benefit of involving stakeholders early in AI policy design?

   • AReduction in the need for technical documentation
   • BFaster policy approval by regulators
   • CGuaranteed elimination of all AI risks
   • DIncreased buy-in and reduced resistance to implementation
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   A healthcare provider is deploying an AI triage system that prioritizes patient urgency. The policy team is integrating risk assessment into the AI governance policy. They have identified potential harms such as mis-triage, bias against certain demographics, and over-reliance by clinicians. Which policy element best addresses these risks?

   • AMandate a pre-deployment risk assessment that includes bias testing, a clinician-in-the-loop override mechanism, and continuous monitoring with defined thresholds for triggering corrective action.
   • BEstablish a policy that allows clinicians to override AI recommendations but does not require any formal risk assessment or monitoring.
   • CDevelop a policy that focuses on legal liability by requiring patients to sign consent forms acknowledging AI involvement.
   • DRequire only a one-time technical validation of the model's accuracy on historical data before deployment.
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   An insurance company is developing an AI underwriting model. The policy team wants to involve stakeholders to ensure the policy addresses fairness and avoids discrimination. Which stakeholder group is most critical to include in the policy design process?

   • AShareholders and investors, because they fund the company and need assurance of compliance.
   • BRegulators only, because they define the legal requirements for fairness.
   • CCustomers and consumer advocacy groups who may be affected by underwriting decisions.
   • DOnly the data science team that builds the model, because they understand the technical details.
   Reveal Solution Discussion Select an answer first