
AAISM (Advanced in AI Security Management) is ISACA's certification for security leaders who govern and secure AI. It is open only to people who already hold an active CISM or CISSP. It covers AI governance and security programmes, AI risk, threats and vendor risk, and AI security architecture, life cycle and data controls. Banks, insurers and GCCs in Hyderabad are rolling out GenAI and agents, and CISOs and security managers there are being asked to answer AI risk questions from boards and regulators. AAISM shows you can.
This guide follows ISACA's official AAISM pages, exam content outline and candidate guide. It covers the exam facts, the three domains and their weightings, key AI threats and controls, a 6-week plan and 15 original practice questions.
Quick facts (ISACA, November 2026)
| Item | Detail |
|---|---|
| Eligibility | An active CISM or CISSP, plus experience in security or advisory roles and with AI systems |
| Exam | 90 multiple-choice questions in 2.5 hours (150 minutes) |
| Fees | US$459 for ISACA members, US$599 for non-members |
| Delivery | PSI test centres or remote proctoring; candidates in India must use test centres |
| Eligibility window | 6 months from registration (one paid extension allowed) |
| Results | Preliminary pass/fail on screen; official result within 10 working days |
| Upcoming change | The exam content outline will be updated from 18 May 2027 |
Read the official AAISM page and exam content outline.
Domains and weightings
| Domain | Weight | Topics |
|---|---|---|
| 1. AI Governance and Program Management | 31% | Stakeholders, industry frameworks and regulation; AI strategies, policies and procedures; AI asset and data life cycle management; AI security programme development; business continuity and incident response |
| 2. AI Risk Management | 31% | AI risk assessment, thresholds and treatment; AI threat and vulnerability management; AI vendor and supply chain management |
| 3. AI Technologies and Controls | 38% | AI security architecture and design; AI life cycle (model selection, training, validation); data management controls; privacy, ethics, trust and safety controls; security controls and monitoring |
The candidate guide calls Domain 2 "AI Risk and Opportunity Management", and the weight is the same. Your score depends on the total number of correct answers, not on the per-domain weights.
AI threats and matching controls
| Threat | Example controls |
|---|---|
| Prompt injection (direct or through documents and tools) | Treat inputs and retrieved content as untrusted; isolate tools and give them least privilege; filter outputs; require human approval for risky actions |
| Data poisoning | Track data provenance; validate training data; restrict who can change datasets; test the model before release |
| Sensitive data leakage | Classify data; use DLP and access controls on retrieval sources; minimise and mask data; keep data in permitted regions |
| Model theft or inversion | Control API access; set rate limits; monitor for unusual extraction patterns |
| Supply chain risk (models, libraries, vendors) | Vendor assessments; model and software bills of materials; contract security requirements; ongoing monitoring |
| Hallucination and unsafe output | Ground answers in approved sources; evaluate against test sets; content safety filters; human oversight |
Frameworks worth knowing for Domain 1 include the NIST AI Risk Management Framework, ISO/IEC 42001 (AI management systems), the EU AI Act and, for India, the Digital Personal Data Protection (DPDP) Act.
6-week study plan
| Week | Focus |
|---|---|
| 1 | AI basics for security leaders: machine learning, GenAI, LLMs, RAG, agents; how AI changes the attack surface |
| 2 | Governance: charter, roles, policies, acceptable use, frameworks and regulation; AI asset inventory and classification |
| 3 | Risk: AI impact and risk assessments, thresholds, treatment; threat landscape; AI red-teaming and vulnerability management |
| 4 | Vendors and supply chain; incident response, business continuity and DR for AI systems |
| 5 | Controls: AI security architecture, life cycle controls, data controls, privacy, ethics and safety, monitoring and metrics |
| 6 | ISACA practice questions; review the 22 supporting tasks; book a test centre slot |
15 practice questions with answers
These are original questions written for this guide, not exam questions.
1. Which certifications qualify you for AAISM?
An active CISM or CISSP.
2. What should come first when an organisation starts using AI widely?
Governance: a charter with clear roles, policies and an inventory of AI systems.
3. Why keep an AI asset inventory?
You can't assess risk or apply controls to AI systems, models and datasets you don't know about.
4. A business unit wants to use a public GenAI tool with customer data. What is the first step?
A risk and impact assessment against policy, data classification and regulatory requirements.
5. What is a risk threshold?
The level of risk the organisation will accept before treatment is required.
6. Which attack hides instructions in a document that an AI assistant reads?
Indirect prompt injection.
7. Which attack corrupts training data to change model behaviour?
Data poisoning.
8. Name one control for AI vendor risk.
Security requirements in the contract, a vendor assessment before onboarding, or continuous monitoring.
9. What does AI red-teaming do?
It tests AI systems with adversarial inputs to find unsafe or insecure behaviour before attackers do.
10. What is human oversight in AI security?
Risk-based review of AI inputs, outputs or actions by people, especially for high-impact decisions.
11. Why include AI systems in business continuity planning?
Critical processes may depend on them, so you need fallbacks if a model or provider becomes unavailable.
12. Which ISO standard covers AI management systems?
ISO/IEC 42001.
13. Which life cycle stage checks that a model meets accuracy, fairness and security requirements before release?
Validation.
14. Give one AI security metric.
For example the number of AI systems assessed, blocked prompt-injection attempts, or time to resolve AI incidents.
15. What should AI acceptable use guidelines tell staff?
Which AI tools are approved, which data may be used with them, and how to report problems.
Before and after AAISM
- You need CISM or CISSP first.
- Add risk and privacy with CRISC and CDPSE; for hands-on AI security, see SecAI+ and SC-500.
- Learn it hands-on in our Cyber Security course and AI course; see live security leadership and GRC jobs.
Cloud Soft Solutions in Ameerpet, Hyderabad teaches cyber security, cloud security and AI 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 AAISM?
ISACA's Advanced in AI Security Management certification, for security leaders who govern and secure AI.
Who can take AAISM?
Candidates with an active CISM or CISSP.
How long is the AAISM exam?
90 multiple-choice questions in 2.5 hours (150 minutes).
What does AAISM cost?
US$459 for ISACA members and US$599 for non-members.
What are the AAISM domains?
AI Governance and Program Management 31%, AI Risk Management 31%, AI Technologies and Controls 38%.
Can I take AAISM online from India?
No. ISACA says candidates in India must take it at a test centre.



