AI Governance and Frameworks CRAGE Practice Questions (Page 4)

Questions 16–20

1. 16application · medium

   A government agency has implemented an AI system for processing visa applications. The agency is required to comply with the EU AI Act and must demonstrate compliance through audits. The agency has appointed a compliance officer but lacks a formal audit process. What should the agency do to ensure ongoing compliance?

   • ARely on the compliance officer's judgment to ensure compliance
   • BConduct an annual internal audit of the AI system and document findings
   • CObtain ISO/IEC 42001 certification instead of conducting audits
   • DPublish the AI system's source code to the public
   Reveal Solution Discussion Select an answer first
2. 17foundation · easy

   Which AI governance framework is a legally binding regulation that categorizes AI applications by risk level and imposes obligations on providers and deployers within a specific jurisdiction?

   • AEU AI Act
   • BNIST AI RMF
   • COECD AI Principles
   • DISO/IEC 42001
   Reveal Solution Discussion Select an answer first
4. 18application · medium

   A financial services firm is integrating an AI system that detects fraudulent transactions. The enterprise risk team wants to include AI-specific risks in the firm's existing risk register. The AI system's false-positive rate could cause customer dissatisfaction and regulatory scrutiny. Which approach best integrates AI risk into the enterprise risk management process?

   • AIgnore the false-positive risk because the system is more accurate than manual review.
   • BDocument the risk in a separate AI-only risk register that is not linked to the enterprise register.
   • CAdd a new risk category for AI-specific risks, including false positives, and assign a risk owner from the AI governance committee.
   • DTreat the AI system as a standard IT system and use the existing IT risk category.
   Reveal Solution Discussion Select an answer first
5. 19expert · medium

   A large enterprise is rolling out an AI governance framework across multiple business units. The data science teams in different units have varying levels of maturity and use different tools. The central governance office wants consistent compliance but also wants to avoid slowing down innovation. Which implementation strategy best balances consistency and flexibility?

   • ADefine a common set of mandatory governance requirements and allow each business unit to implement them using their own tools and processes.
   • BRequire all business units to use the same centralized AI development platform and approval workflow.
   • CLet each business unit develop its own governance framework independently.
   • DImplement the framework only in the most mature business unit and expand later.
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   A retail company wants to implement an AI chatbot for customer service. The company's leadership values transparency and fairness. The legal team warns that the chatbot must not discriminate against protected groups. Which policy element should the company include in its AI governance policy to address these concerns most directly?

   • AMandate that all chatbot responses be reviewed by a human before sending.
   • BRequire regular bias audits and publish a plain-language explanation of how the chatbot makes decisions.
   • CLimit the chatbot to answering only pre-approved questions from a fixed list.
   • DStore all chatbot logs for five years to enable future investigations.
   Reveal Solution Discussion Select an answer first