How to pass the AIGP
A practical guide to passing the IAPP Artificial Intelligence Governance Professional (AIGP) exam: what it tests, how to study, and where people lose marks. Written by someone who sat it and scored well.
What the AIGP actually tests
The AIGP certifies that you can help an organisation build and run responsible AI governance. It is broad, spanning technology, ethics, risk and law. The IAPP body of knowledge maps to eight areas, all covered free in the study notes:
- Foundations of AI: what AI is, the intelligence ladder (ANI to ASI), the OECD classification, and how machines learn.
- AI impacts and responsible AI: harms and bias taxonomies, the OECD principles, and trustworthy AI.
- Governance and risk management: roles, governance structures, risk frameworks and standards such as the NIST AI RMF and ISO/IEC 42001.
- AI regulation: the EU AI Act risk tiers and obligations, and emerging laws worldwide.
- How existing laws apply to AI: privacy, intellectual property, liability and anti-discrimination.
- Governing AI development: the AI life cycle, planning, data, testing and documentation.
- Governing AI deployment: monitoring, post-market oversight and incident response.
- The vocabulary every candidate is expected to recognise on sight.
The exam format
The AIGP is a multiple-choice exam of 100 questions. The official practice exam is built to the same blueprint, and the IAPP suggests timing yourself at three hours to simulate exam conditions. Confirm the current format and scoring on the IAPP's official exam blueprint before you book.
A study plan that works
- Weeks 1 to 2, build the map. Read the notes module by module so you can place every concept, especially the EU AI Act structure and the four roles (provider, deployer, importer, distributor).
- Weeks 3 to 4, drill. Move to questions. After each module, answer practice questions and read every explanation, including the ones you got right.
- Week 5, spaced review. Let the spaced-repetition schedule resurface weak spots across all eight areas.
- Final week, simulate. Sit the full official practice exam under timed conditions and review every miss.
Where people lose marks
- ANI versus AGI versus ASI. If a scenario describes a deployed real-world system, the answer is narrow AI; AGI and ASI do not yet exist.
- The EU AI Act. Know the risk tiers (unacceptable, high, limited, minimal) and which obligations attach to which role.
- Bias and harms taxonomies. Be able to tell computational, cognitive and societal bias apart, and individual from group and societal harms.
- Frameworks. Do not confuse the NIST AI RMF, the OECD principles and ISO/IEC 42001; know what each is for.
- Reading too fast. Many items turn on a single qualifying word. Slow down on the stem.
How this site helps
The study notes are free and structured for active recall, with the exam-critical wording highlighted. The practice question bank includes the official practice exam plus the module question sets, each with a worked explanation, marked automatically and scheduled for spaced review. Your progress syncs across your devices.