
AAIA (Advanced in AI Audit) is ISACA's certification for auditors who give assurance over AI. It is for IT audit professionals who already hold a CISA, or a qualifying audit or accounting credential such as CIA, US CPA or ACCA together with an IT audit or advisory role. It covers AI governance and risk, AI operations (data, development, change, testing, threats and incidents) and AI audit tools and techniques. Big-4 firms, banks and GCCs in Hyderabad are building AI audit teams, and auditors who can test AI controls are in short supply.
This guide follows ISACA's official AAIA exam content outline and the 2026 candidate guide. It covers the exam facts and eligibility, the three domains and their weightings, an AI audit work programme, a 6-week plan and 15 original practice questions.
Quick facts (ISACA, November 2026)
| Item | Detail |
|---|---|
| Eligibility | CISA holders qualify. Holders of CIA, US CPA, ACCA, ICAEW ACA and other listed credentials qualify if they work in IT audit or advisory |
| Exam | 90 multiple-choice questions in 2.5 hours (150 minutes) |
| Languages | English, Simplified Chinese, Spanish, Japanese |
| Fees | US$459 for ISACA members, US$599 for non-members |
| Scoring | Based on the total number of correct answers; domain results are for information only |
Read the official AAIA exam content outline. ISACA requires candidates in India to sit its AAISM exam at a test centre, so check the AAIA delivery options for India when you register.
Domains and weightings
| Domain | Weight | Topics |
|---|---|---|
| 1. AI Governance and Risk | 33% | AI models, considerations and requirements; AI governance and programme management; AI risk management; privacy and data governance programmes; leading practices, ethics, regulations and standards |
| 2. AI Operations | 46% | Data management for AI; development methodologies and life cycle; change management; supervision of AI outputs, impacts and decisions; testing techniques; AI threats and vulnerabilities; AI incident response |
| 3. AI Auditing Tools and Techniques | 21% | Audit planning and design; testing and sampling; evidence collection; data quality and analytics; AI audit outputs and reports |
AI Operations is almost half the exam. Expect questions on how to test controls around data, models, changes and monitoring.
An AI audit work programme in brief
| Area | What to test | Evidence |
|---|---|---|
| Governance | AI policy, ownership, inventory of AI systems, risk appetite | Policies, charter, AI register, committee minutes |
| Data | Lineage, quality, consent and privacy, bias checks | Data catalogue, quality reports, data protection impact assessments |
| Development | Requirements, validation, independent review before release | Model cards, validation reports, approval records |
| Change management | Retraining and prompt or model changes go through approval and testing | Change tickets, version history, test results |
| Supervision | Human oversight, monitoring of drift and output quality | Monitoring dashboards, override logs, escalation records |
| Security | Prompt injection, data leakage, access to models and APIs | Red-team results, access reviews, security test reports |
| Incidents | AI incidents are logged, investigated and fed back into controls | Incident register, root-cause analyses |
Frameworks worth knowing include the NIST AI Risk Management Framework, ISO/IEC 42001, the EU AI Act and, for India, the Digital Personal Data Protection (DPDP) Act.
6-week study plan
| Week | Focus |
|---|---|
| 1 | AI and ML basics for auditors: model types, GenAI, training vs inference, RAG and agents |
| 2 | Governance, risk and privacy; ethics, regulations and standards |
| 3 | Data management for AI; development life cycle; change management |
| 4 | Supervision of AI outputs; testing techniques; threats, vulnerabilities and incident response |
| 5 | Audit planning, sampling, evidence, data analytics and reporting for AI |
| 6 | ISACA practice questions, weak-area review; book the exam |
15 practice questions with answers
These are original questions written for this guide, not exam questions.
1. Who qualifies for AAIA without further conditions?
CISA holders.
2. What should an auditor check first when an organisation uses AI?
Whether there is governance: an AI policy, clear ownership and an inventory of AI systems.
3. What is model drift?
A drop in model performance over time as real-world data changes from the training data.
4. Which control finds drift?
Ongoing monitoring of model performance against agreed thresholds.
5. Why audit training data lineage?
To confirm where data came from, that it may be used, and that it hasn't been tampered with.
6. A model was retrained without approval. Which control failed?
Change management.
7. What does human oversight mean for an AI credit decision?
A person can review, override and be accountable for high-impact decisions.
8. Name one testing technique for AI solutions.
For example validation against a held-out test set, bias testing, red-teaming or adversarial testing.
9. Which AI threat inserts malicious instructions through user input?
Prompt injection.
10. Why is sampling harder for AI outputs?
Outputs can vary for similar inputs, so samples must be designed to cover risk scenarios, not just random volume.
11. What makes audit evidence reliable?
It is relevant, sufficient, obtained independently where possible, and documented.
12. How can data analytics help an AI audit?
By testing full populations, for example comparing all AI decisions against policy rules.
13. Which ISO standard sets requirements for an AI management system?
ISO/IEC 42001.
14. What should an AI audit report include?
Scope, findings with risk ratings, root causes, recommendations and management responses.
15. An AI system caused customer harm. What should the incident process produce?
A logged incident, an investigation and root-cause analysis, and control improvements.
Before and after AAIA
- Start with CISA; add risk with CRISC and governance with CGEIT.
- For the security side of AI, see AAISM; for privacy, see CDPSE.
- Learn AI hands-on in our AI, GenAI and Agentic AI course and security in our Cyber Security course; see live audit, risk and GRC jobs.
Cloud Soft Solutions in Ameerpet, Hyderabad teaches AI, cloud and cyber security with real-time projects. See the APEX program and all courses.
📞 Call +91 96660 19191 · WhatsApp · Flat No. 513, 5th Floor, Nilagiri Block, Aditya Enclave, Ameerpet, Hyderabad
Frequently Asked Questions
What is AAIA?
ISACA's Advanced in AI Audit certification, for auditors who provide assurance over AI.
Who can take AAIA?
CISA holders, or holders of listed credentials such as CIA, US CPA or ACCA who work in an IT audit or advisory role.
How long is the AAIA exam?
90 multiple-choice questions in 2.5 hours (150 minutes).
What are the AAIA domains?
AI Governance and Risk 33%, AI Operations 46%, AI Auditing Tools and Techniques 21%.
What does AAIA cost?
US$459 for ISACA members and US$599 for non-members.
Which languages is AAIA offered in?
English, Simplified Chinese, Spanish and Japanese.



