ISACA · AAIA

Advanced in AI Audit practice questions

IT auditors who audit AI systems

Pass markScaled score; see ISACA
Questions in the exam90
Practice questions here25

What the AAIA exam covers

Current exam outline
AI governance and risk33%
AI operations46%
AI auditing tools and techniques21%

Sample AAIA questions

Five of the 25 practice questions, each with its answer and explanation.

AI governance and risk

What should an auditor first confirm about an organisation's use of AI?

  1. The number of developers
  2. That a complete inventory of AI systems exists, with owners
  3. The marketing budget
  4. The model's colour scheme
Show the answer

B. That a complete inventory of AI systems exists, with owners

An inventory defines the audit universe.

AI governance and risk

What does an organisation's AI risk appetite set out?

  1. The size of its servers
  2. How much AI-related risk it is willing to accept in pursuit of its goals
  3. Its holiday schedule
  4. Its logo guidelines
Show the answer

B. How much AI-related risk it is willing to accept in pursuit of its goals

Risk appetite guides decisions on which AI uses are acceptable.

AI governance and risk

An AI system processes personal data in a new way. What assessment is typically required?

  1. None
  2. A data protection impact assessment
  3. A cost-benefit analysis only
  4. A penetration test only
Show the answer

B. A data protection impact assessment

Privacy impact assessments evaluate risks to individuals.

AI governance and risk

Which can serve as criteria for auditing an AI management system?

  1. A product brochure
  2. A press release
  3. Social media posts
  4. ISO/IEC 42001 and the NIST AI RMF
Show the answer

D. ISO/IEC 42001 and the NIST AI RMF

Recognised standards and frameworks provide objective criteria.

AI governance and risk

A bank's AI model declines loans. What requirement is especially relevant?

  1. Faster response times only
  2. Hiding the use of AI
  3. Being able to explain decisions to affected customers and regulators
  4. Using the largest model
Show the answer

C. Being able to explain decisions to affected customers and regulators

Explainability supports fairness, accountability and compliance.