AAIA

What should you ask in an AAIA interview, and how should candidates answer?

An AAIA holder should be able to audit an AI system on evidence: what it was approved to do, what it does now, which version is running, and who checks its outputs. These questions follow the three exam domains, with what a strong answer covers. Candidates get the questions to prepare for and four to ask back.

What should an employer ask an AAIA candidate?

Each question maps to an exam domain, so you test what the role needs. Score each answer 1 to 4 on the free scorecard below.

  1. A vendor reports one accuracy figure for the model we are buying. What do you ask next?

    Listen for: Accuracy on what data and for which version; how rare the outcome is; recall and precision where the two errors cost different amounts; results by group; and whether the figure says anything about our own deployment.

    Domain 1: AI Governance and Risk (AI models, considerations and requirements)

  2. How would you test whether our AI inventory is complete?

    Listen for: Treat it as a claim. Reconcile it to sources the AI program does not control: purchase orders, expense claims, sign-in records, cloud bills and vendor release notes. A system found outside it is a finding, and so is the control that missed it.

    Domain 1: AI Governance and Risk (AI governance and program management)

  3. A supplier tells us it is ISO/IEC 42001 certified. How much do you rely on that?

    Listen for: Check accreditation, scope and dates, then read the statement of applicability for exclusions. The certificate covers a management system on a sample, not a model's accuracy or fairness, and it is not proof of EU AI Act compliance.

    Domain 1: AI Governance and Risk (leading practices, ethics, regulations and standards)

  4. How would you audit change management for a model that retrains every week?

    Listen for: Get the version history first. Test the controls on the pipeline: bounds on what a retrain may change, tests before promotion, monitoring after each release, rollback. For hosted models, version pinning and notice in the contract.

    Domain 2: AI Operations (change management specific to AI)

  5. Our claims team says a person reviews every model decision. How would you test that?

    Listen for: Watch the work. Override rates, review time per case, seeded cases with a known right answer, and whether a reviewer can disagree without penalty. A signed procedure shows only that oversight was designed.

    Domain 2: AI Operations (supervision of AI solutions)

  6. What AI-specific threats would you expect in the risk register for a customer assistant built on a language model?

    Listen for: Prompt injection, data leakage, poisoning, insecure output handling and excessive agency, each tied to a control that touches the model, its data or what it can do. A single line saying cyber attack is a finding about the risk assessment.

    Domain 2: AI Operations (threats and vulnerabilities specific to AI)

  7. Halfway through fieldwork you find a model wrongly refusing applicants, and it is still happening. What do you do?

    Listen for: Tell management and the incident owner at once. Do not switch it off yourself: that is a management decision. Preserve the evidence, and report it later as a finding on detection, containment and root cause.

    Domain 2: AI Operations (incident response management specific to AI)

  8. The model you sampled was retrained halfway through the audit period. What does that do to your conclusion?

    Listen for: Each version is a stratum. Sample from each or limit the conclusion to the versions tested and say so. Test the change control over the transition, and roll forward from an interim conclusion.

    Domain 3: AI Auditing Tools and Techniques (audit testing and sampling)

  9. Write the condition for a fairness finding on a claims model.

    Listen for: Stated as fact a reader can check: model version, period, the measured gap, the approved tolerance and the population tested. Criterion, cause, effect and recommendation follow.

    Domain 3: AI Auditing Tools and Techniques (AI audit outputs and reports)

  10. How would you use an AI tool in your own audit work, and what would you never let it do?

    Listen for: An approved tool whose terms rule out retention and training on inputs. Every factual output traced to its source before it enters a workpaper. Tool, version, prompt and output recorded. A sample of its work reviewed. The auditor signs the report, not the tool.

    Domain 3: AI Auditing Tools and Techniques (AI within the audit process)

AAIA interview scorecard

AAIA interview scorecard

Candidate: ______ Interviewer: ______ Date: ______

#QuestionListen forScore 1 to 4
1A vendor reports one accuracy figure for the model we are buying. What do you ask next?Accuracy on what data and for which version; how rare the outcome is; recall and precision where the two errors cost different amounts; results by group; and whether the figure says anything about our own deployment.
2How would you test whether our AI inventory is complete?Treat it as a claim. Reconcile it to sources the AI program does not control: purchase orders, expense claims, sign-in records, cloud bills and vendor release notes. A system found outside it is a finding, and so is the control that missed it.
3A supplier tells us it is ISO/IEC 42001 certified. How much do you rely on that?Check accreditation, scope and dates, then read the statement of applicability for exclusions. The certificate covers a management system on a sample, not a model's accuracy or fairness, and it is not proof of EU AI Act compliance.
4How would you audit change management for a model that retrains every week?Get the version history first. Test the controls on the pipeline: bounds on what a retrain may change, tests before promotion, monitoring after each release, rollback. For hosted models, version pinning and notice in the contract.
5Our claims team says a person reviews every model decision. How would you test that?Watch the work. Override rates, review time per case, seeded cases with a known right answer, and whether a reviewer can disagree without penalty. A signed procedure shows only that oversight was designed.
6What AI-specific threats would you expect in the risk register for a customer assistant built on a language model?Prompt injection, data leakage, poisoning, insecure output handling and excessive agency, each tied to a control that touches the model, its data or what it can do. A single line saying cyber attack is a finding about the risk assessment.
7Halfway through fieldwork you find a model wrongly refusing applicants, and it is still happening. What do you do?Tell management and the incident owner at once. Do not switch it off yourself: that is a management decision. Preserve the evidence, and report it later as a finding on detection, containment and root cause.
8The model you sampled was retrained halfway through the audit period. What does that do to your conclusion?Each version is a stratum. Sample from each or limit the conclusion to the versions tested and say so. Test the change control over the transition, and roll forward from an interim conclusion.
9Write the condition for a fairness finding on a claims model.Stated as fact a reader can check: model version, period, the measured gap, the approved tolerance and the population tested. Criterion, cause, effect and recommendation follow.
10How would you use an AI tool in your own audit work, and what would you never let it do?An approved tool whose terms rule out retention and training on inputs. Every factual output traced to its source before it enters a workpaper. Tool, version, prompt and output recorded. A sample of its work reviewed. The auditor signs the report, not the tool.

Source: https://credentialpress.com/guides/aaia-interview-questions

Which questions should an AAIA candidate prepare for?

  1. Why the AAIA on top of the credential you already hold?

    How to answer: Name the AI work that reached your desk and the point where your audit method bent. One engagement is enough.

  2. Which domain do you know best, and which least?

    How to answer: AI Operations is 46% of the exam and most of the job. Give one example from your strongest, then say how you are closing the gap on the weakest.

  3. Which AI system here would you look at first, and why?

    How to answer: Rank by effect on people, autonomy, scale and rate of change, and say what evidence you would ask for in week one.

  4. How do you stay independent when the data scientists know the model better than you?

    How to answer: Bring in an expert where the evidence is beyond you, keep the conclusion yours, never test a model with the auditee's own tool, and never audit a control you helped design.

What should a candidate ask the employer?

  • Do you have an AI inventory, and how do you know it is complete?
  • Which AI systems here make or shape decisions about people?
  • Can the audit team get read access to model version history, logs and monitoring, or will we work from screenshots?
  • Which AI tools is the audit team approved to use on auditee data?

Where next?

Also free: the AAIA mind map and the AAIA 10-question quiz, plus every other credential on our study tools page. For the full syllabus, the AAIA Exam Guide and the AAIA Practice Questions go domain by domain.

Frequently asked questions

What should an employer ask an AAIA candidate?

Questions that test each exam domain in practice, for example: A vendor reports one accuracy figure for the model we are buying. What do you ask next? How would you test whether our AI inventory is complete? A supplier tells us it is ISO/IEC 42001 certified. How much do you rely on that?

What should an AAIA candidate ask the employer?

Do you have an AI inventory, and how do you know it is complete? Which AI systems here make or shape decisions about people? Can the audit team get read access to model version history, logs and monitoring, or will we work from screenshots? Which AI tools is the audit team approved to use on auditee data?

How should a candidate prepare for an AAIA interview?

Name the AI work that reached your desk and the point where your audit method bent. One engagement is enough.

Which books go deeper on AAIA?

Cover of AAIA Exam Guide

AAIA Exam Guide

A study companion for the ISACA Advanced in AI Audit (AAIA) examination. 13 chapters, 371 pages.

Cover of AAIA Practice Questions

AAIA Practice Questions

Practice questions with full rationales, sized to the published domain weights. 216 pages.

How to credit this page

Using a figure or a passage from this page in an article, a course, a forum answer or AI-generated text? Please credit Credential Press and link to this page, so your readers can check the source and the date it was reviewed. Copy a ready-made citation:

Plain text

Markdown

HTML