Why Study with PlanetCert?
The Latest Questions
Practice questions and exam topics aligned with the current exam objectives.
Detailed Explanations
Go beyond the answer. Master the material with comprehensive learning and professional explanations for every concept.

AI-Powered Insights
Personalized preparation guidance that adapts to your performance and identifies weak spots automatically.
Exam Information
Official specifications published by ISACA
Exam Format
Registration
Validity
Skills this practice test covers
- Set up an AI governance charter with roles that align AI use with business objectives.
- Map AI systems to regulations and frameworks such as the EU AI Act, GDPR and ISO/IEC 42001.
- Assess an AI risk, set its threshold against the risk appetite and choose a treatment.
- Spot adversarial attacks such as evasion, poisoning and extraction, and pick the control for each.
- Check an AI vendor through due diligence, contract terms and ongoing monitoring.
- Protect training data and models with data controls, access controls and adversarial testing.
Explore related topics
AAISM exam content
3 domains · 24 topics · 226 questions
AI Governance and Program Management31%70 questions
This domain decides who owns AI security and under which rules. Candidates help write an AI governance charter, align AI use with laws and frameworks, and run an AI security program with policies, an asset inventory, training and incident response.
Skills tested
- Draft a governance charter with roles and responsibilities for AI.
- Align AI use with regulations and frameworks such as the EU AI Act and ISO/IEC 42001.
- Write AI-specific security policies, including an AI acceptable use policy.
- Inventory and classify AI assets, data and models across their life cycles.
- Plan incident response and business continuity for AI systems.
Questions by topic
- Stakeholder Considerations, Industry Frameworks, and Regulatory Requirements11
- AI-Related Strategies, Policies, and Procedures5
- AI Asset and Data Life Cycle Management13
- Business Continuity and Incident Response6
- AI Security Program Development and Management10
- Industry Frameworks and Standards5
- Regulatory and Legal Requirements9
- Privacy, Ethical, Trust and Safety Controls2
- Data Management Controls2
- Security Controls and Monitoring1
- AI Security Architecture and Design1
- AI Risk Assessment, Thresholds, and Treatment2
- AI Vendor and Supply Chain Management3
AI Risk Management31%66 questions
This domain applies the risk life cycle to AI. Candidates identify and analyze AI risks, set thresholds that fit the risk appetite, choose a treatment, and manage AI threats, vulnerabilities and vendor risk.
Skills tested
- Identify AI-specific risks and judge their effect on enterprise risk.
- Set risk thresholds that match the risk appetite, then choose a treatment.
- Track AI threats and vulnerabilities, including adversarial attacks on models.
- Assess AI vendors through due diligence, contract terms, service level agreements and continuous monitoring.
- Map the AI supply chain, including outside models and their dependencies.
Questions by topic
- AI Risk Assessment, Thresholds, and Treatment24
- AI Threat and Vulnerability Management32
- AI Vendor and Supply Chain Management8
- Privacy, Ethical, Trust and Safety Controls2
AI Technologies and Controls38%90 questions
The largest domain is about the controls themselves. Candidates design secure AI architectures, protect the model life cycle from selection to monitoring, control data, and apply privacy, ethics, trust and safety measures such as bias testing and explainability.
Skills tested
- Design a secure AI architecture with defense-in-depth across models, data and infrastructure.
- Secure model selection, training, validation and deployment, with adversarial testing before release.
- Apply data controls, such as classification, validation and sanitization, throughout the AI data life cycle.
- Choose privacy-preserving techniques such as differential privacy and federated learning.
- Add bias testing and human-in-the-loop review where AI outputs affect people.
Questions by topic
- Privacy, Ethical, Trust and Safety Controls33
- AI Security Architecture and Design9
- AI Life Cycle (Model Selection, Training, and Validation)17
- Security Controls and Monitoring18
- Data Management Controls11
- AI Threat and Vulnerability Management1
- AI Asset and Data Life Cycle Management1
Requirements
- Required: an active CISM or CISSP certification.
- Recommended: hands-on work assessing, implementing and maintaining AI systems, from a security or advisory role.
- Recommended: knowledge of enterprise risk management, and familiarity with AI technologies and architectures.
About the AAISM exam
AI brings new assets, new attacks and new rules into a security program, and ISACA's AAISM exam tests whether a security manager can govern all three. It requires an active CISM or CISSP certification, so it builds on security management experience instead of teaching it. Of its three domains, AI Technologies and Controls carries the most weight; the other two cover AI governance and program management, and AI risk management. Passing earns the ISACA Advanced in AI Security Management certification.
Questions in the PlanetCert practice test lean toward AI Technologies and Controls, especially privacy, ethics, trust and safety controls. Within AI Risk Management, most questions deal with AI threats and vulnerabilities. The explanations link each answer to a policy, a risk decision or a control. Study Mode suits the risk questions, and Brainy AI Assistant can explain an attack such as model extraction and the controls that limit it.
How to earn the certification
- Earns
- ISACA Advanced in AI Security Management (AAISM)Advanced
AAISM is an ongoing certification with annual requirements that holders must meet.
- Before
- CISM Certified Information Security Manager (CISM)
One of the two credentials that qualify a candidate; an active CISM meets the requirement.
- CISM Certified Information Security Manager (CISM)
How to prepare for AAISM
- 1
Note where AI changes security work
AAISM builds on security management skills that candidates already have. For each one, note what AI adds, such as models as new assets, training data as a new attack target and model vendors as new suppliers.
- 2
Match attacks to their controls
Write each adversarial attack type next to the control that limits it. For example, data validation limits poisoning, and limits on queries to a model's API make extraction harder.
- 3
Give controls the most study time
AI Technologies and Controls is the largest domain. Work through secure architecture, the model life cycle, data controls, and privacy techniques such as differential privacy.
- 4
Map one AI system to its rules
Pick one AI system, such as a customer chatbot, and list the laws, frameworks and policies that apply to it. Then write its risk entry with a threshold and a treatment.
- 5
Finish on timed mixed-domain practice
Take the AAISM practice test in Timed Mode, then read the explanation for every miss. Ask Brainy AI Assistant to compare two controls when both seem to fit a scenario.
Start a timed AAISMrun
Who AAISM is for
- Security managers with an active CISM or CISSP who now answer for AI systems.
- Chief information security officers and security directors who set AI policy.
- Governance, risk and compliance staff who must fit AI into existing controls.
- IT auditors who review AI systems and want the security manager's view.
Typical job titles: Chief information security officer · Security manager · AI security specialist · Risk manager · GRC manager
AAISM glossary
17 terms from the exam objectives and the vendor's official material, each defined in a sentence.
- Adversarial attacks (evasion, poisoning, extraction)
- Attacks that fool a model with crafted inputs (evasion), corrupt its training data (poisoning) or copy it through its outputs (extraction).
- AI Model Registries
- Central records of models, with their versions, owners and approval status.
- AI steering committee
- A group of senior leaders who set the direction for AI use and decide on major AI risks.
- Business impact analysis
- A study of how losing a system or process would hurt the business, used to set recovery priorities.
- COBIT 2019
- ISACA's framework for the governance and management of enterprise IT, which can be applied to AI.
- Data anonymization and pseudonymization
- Anonymization removes identifying details for good; pseudonymization replaces them with codes that can be linked back under control.
- Defense-in-depth for AI systems
- Layering independent controls around an AI system, so that one failed control does not expose it.
- Differential privacy
- A technique that adds controlled noise to data or results, so that no single person's record can be picked out.
- Due diligence
- Checks on a vendor's security, practices and record before a contract is signed.
- EU AI Act
- The European Union law that sets rules for AI systems according to their level of risk.
- Explainable AI (XAI) methods
- Techniques that show why a model produced an output, so that people can check and trust the result.
- Federated learning
- Training one model across many devices or sites without collecting the raw data in one place.
- GDPR
- The European Union's data protection regulation, which also applies when AI systems process personal data.
- Homomorphic encryption
- Encryption that lets a system compute on data while the data stays encrypted.
- Human-in-the-loop controls
- Controls in which a person reviews or approves an AI decision before it takes effect.
- ISO/IEC 42001
- The international standard for an AI management system, which an organization uses to govern its AI.
- Risk appetite
- The amount and kind of risk an organization is willing to accept to meet its goals.
Related practice tests
- CCOACertified Cybersecurity Operations Analyst
- CRISCCertified in Risk and Information Systems Control (CRISC)
- CGEITCertified in the Governance of Enterprise IT (CGEIT)
- CISMCertified Information Security Manager (CISM)
- CISACertified Information Systems Auditor (CISA)