
Course introduction: what the ISACA AAIR exam prep covers across its three domains, from AI governance frameworks to risk methods and regulation, and how to use the course.
Why AI needs a dedicated governance response as it moves from experiment to enterprise-critical system, and what that means for risk professionals.
How the AAIR exam is structured around its three weighted domains, governance, lifecycle risk and program management, and how to plan your study and exam strategy.
A technical primer on AI for risk practitioners, covering the core vocabulary and the AI lifecycle needed to describe systems and assess their risk.
How AI risks are classified by category and source, including strategic, operational, compliance, reputational and financial risk, and where AI-specific risks arise.
How the NIST AI Risk Management Framework 1.0 works, with Govern as the cross-cutting function and Map, Measure and Manage around it.
How ISO/IEC 42001, the first certifiable AI management system standard, gives organisations an auditable structure for managing AI risk across the lifecycle.
How ISO/IEC 23894 applies the ISO 31000 risk management process specifically to AI, and how it complements ISO/IEC 42001.
How the EU AI Act classifies AI systems into risk tiers, the obligations at each tier, and how compliance duties differ between providers and deployers.
How sector rules in finance, healthcare and critical infrastructure build on horizontal frameworks such as the NIST AI RMF and the EU AI Act.
How ISACA's Digital Trust Ecosystem Framework turns digital trust principles into practical AI governance structures, roles and oversight.
How to compare and combine NIST AI RMF, ISO/IEC 42001, ISO/IEC 23894 and the EU AI Act, a key skill tested in the AAIR exam.
A risk map of the AI lifecycle, showing where risk concentrates from problem framing and data through deployment to retirement.
How data quality, bias and lineage create AI risk, and how to assess whether training data meets regulatory and operational needs.
How AI models drift and degrade over time, why independent validation is essential, and what performance monitoring the major frameworks expect.
How bias, fairness and explainability risks arise in AI systems, the regulatory and reputational harm they cause, and the controls that reduce them.
How the OWASP Top 10 for LLM applications classifies the most critical security risks, such as prompt injection, and how to use it in threat modelling.
How MITRE ATLAS catalogues adversarial tactics and techniques against machine learning systems, and how to use it to model AI threats.
How AI red teaming and adversarial testing probe model behaviour and safety before production, and how they differ from conventional security testing.
How to manage the risks of third-party foundation models, AI APIs and embedded vendor AI that normal vendor management was not built for.
How to detect, respond to and learn from AI incidents, and why they need governed processes beyond normal IT incident management.
How to design, scope and run an enterprise AI risk program that links board oversight to day-to-day risk work.
A structured, repeatable method to identify, analyse, evaluate and treat AI risks, building on enterprise risk management principles.
How preventive, detective and corrective controls map across the AI lifecycle, from data sourcing and model development to deployment and monitoring.
How to design AI key risk indicators, set thresholds and escalation routes, and build continuous AI risk monitoring.
How to report AI risk to the board in business language, structure reporting lines and frame AI risk appetite for non-technical leaders.
How internal audit gives independent assurance over AI, assessing controls and governance while keeping its independence.
How to fit AI risk into existing enterprise and IT risk frameworks rather than leaving it to data science teams alone.
Why agentic AI systems that plan and act on their own create new risk categories, and how governance frameworks must adapt.
How quantum computing threatens the cryptography protecting model weights, training data and inference pipelines, and how post-quantum migration applies to AI security.
How AI regulation is developing beyond the EU AI Act across major jurisdictions, and what it means for enterprise compliance planning.
Exam practice for Domain 1, governance and frameworks: high-frequency topics, elimination logic and the governance mindset the exam rewards.
Exam practice for Domain 2, AI lifecycle risk, covering risks at each stage from data acquisition to model validation and monitoring.
Exam practice for Domain 3, AI risk program management: AI inventory, risk appetite, three lines of defence, key risk indicators and board reporting.
A timed practice set covering all three AAIR domains to test exam readiness, followed by a course wrap-up.
This course contains the use of artificial intelligence.
AI risk has moved from the IT department to the boardroom. The EU AI Act is in force, the NIST AI Risk Management Framework is cited in procurement, and clients are asking for ISO 42001.
ISACA's Advanced in AI Risk (AAIR) credential is aimed at credentialled IT risk professionals who now have to govern AI. It is a 90-question, scenario-based exam that tests applied judgement, not definitions.
This course prepares you for all three exam domains: AI Risk Governance and Framework Integration, AI Lifecycle Risk Management, and AI Risk Program Management.
WHAT YOU WILL LEARN TO DO
By the end of this course you will be able to:
HOW THE COURSE IS TAUGHT
WHAT IS COVERED
100 PRACTICE QUESTIONS
Every section ends with a quiz built around realistic scenarios, so you test your understanding as you go and reinforce what you have learned. Each answer comes with a short explanation.
HANDS-ON PRACTICE
Enrol today and sit AAIR with the frameworks and the exam technique in place.