The IAPP tells you how long to study for the AIGP and in what order, and most candidates never read it. Its study guide says to "plan for a minimum of 30 hours of study time in advance of your exam date" and lays out four steps, starting with the Body of Knowledge. What it does not do is turn those 30 hours into a schedule, or say which of the four domains deserves the most of them. The exam blueprint does: it gives every domain and every competency a range of scored questions. This guide puts the two together into a plan you can start this week, with the IAPP's own rules on what to study from, how the questions are pitched, and what to do in the last days before the exam.
The IAPP's four steps
The study guide's advice fits on one page. "In general, the IAPP recommends you plan for a minimum of 30 hours of study time in advance of your exam date; however, you might need more or fewer hours depending on your personal choices and professional experience." Then four steps, in this order:
- Review the body of knowledge. "The body of knowledge for the AIGP is a comprehensive outline of the subject matter areas covered by the AIGP exam." The blueprint inside it "indicates the minimum and maximum number of items included on the AIGP exam from the major areas", and "Questions may be asked from any of the topics listed under each area".
- Get certification training. The IAPP's own classes, in person, live online or self-paced. Optional, and the FAQs say so: "No one paid resource is required to pass the exam".
- Take the AIGP practice exam. "The AIGP practice exam consists of 100 questions in the same format as the official certification exam."
- Purchase and schedule your exam. From the IAPP store, to be taken "in person at a Pearson VUE testing center, or remotely via Pearson VUE's OnVUE system."
The handbook adds the standard to judge yourself by. Candidates "should judge exam preparedness by their thorough understanding of all topics on the body of knowledge rather than completion of any given training material", because "Candidates may be asked to apply their knowledge to different situations not outlined in the preparation materials."
Study the document the exam is written from
Step one is the whole plan's foundation, and the IAPP says so three times over. The handbook: "No question will be on the exam that is not covered by a topic or subtopic on that exam's body of knowledge. The body of knowledge should serve as the candidate's core resource." The FAQs: "The exam is created based on this exact document, so nothing will appear on the test that isn't on the body of knowledge."
The Body of Knowledge is four domains, 13 competencies and 58 performance indicators, and the questions are written to the indicators: "Exam questions assess an AI governance professional's proficiency on the performance indicators." So the first hour of any plan is reading all 58, once, as a checklist. The AIGP Body of Knowledge v2.1, explained walks through them; the AIGP exam guide lists every indicator under its competency.
Make sure it is version 2.1, in force for anyone testing on or after February 2, 2026. The exam information page carries both editions and labels the current one "Use if testing on or after 2 February 2026".
Weight the hours by the blueprint
The blueprint gives each domain a range of the 85 scored questions. Taking the midpoint of each range as its share and spreading 30 hours in proportion gives this split, rounded to the half hour.
| Domain | Scored questions | Midpoint | Share of 85 | Hours of 30 |
|---|---|---|---|---|
| Understanding the foundations of AI governance | 16 to 20 | 18 | 21% | 6.5 |
| Understanding how laws, standards and frameworks apply to AI | 19 to 23 | 21 | 25% | 7.5 |
| Understanding how to govern AI development | 21 to 25 | 23 | 27% | 8 |
| Understanding how to govern AI deployment and use | 21 to 25 | 23 | 27% | 8 |
Within the domains, two competencies stand out. IV.C, governing deployment and use, carries 9 to 11 questions, and III.C, release, monitoring and maintenance, carries 8 to 10. Between them that is 17 to 21 scored questions, about as many as all of Domain I. Give them the first and longest sessions of their domains' hours.
The smallest competency, II.D on industry standards and tools, carries 3 to 5 questions on three named sources: the OECD's AI principles, the NIST AI Risk Management Framework and Playbook, and "the core ISO AI standards (i.e., 22989, 42001 and 42005)". It is small enough to learn completely, which is the aim for every low-range competency: three indicators tested up to five times means each one is tested more than once.
A six-week plan: five weeks of content, one of rehearsal
Thirty hours at six a week is five weeks of content, and a sixth week, beyond the minimum, is for the mock and the exam itself. The order follows the Body of Knowledge, which builds from definitions to law to practice, with the heavy domains in the middle.
Week 1, Domain I (6.5 hours). What AI is, the risks and harms it poses, why it needs governance, the principles of responsible AI, roles and responsibilities, and the policies that run across the life cycle. This is the vocabulary the other three domains use. Finish the week by answering the practice questions for its three competencies and reading every rationale, right or wrong.
Week 2, Domain II, part one (4 hours). The laws that already applied before any AI law existed: II.A, how data privacy law's transparency, lawful basis, data minimization, controller obligations and special-category rules apply to AI; and II.B, intellectual property, nondiscrimination, consumer protection and product liability. Learn these as applications, because every indicator begins with "Understand" and the questions ask how a known rule bears on an AI system.
Week 3, Domain II, part two (3.5 hours). II.C, the AI-specific laws, through the EU AI Act's structure of risk classes, requirements, oversight, general-purpose models, enforcement and roles, and then the other laws the domain description names. And II.D, the three standards; the EU AI Act, the NIST AI RMF and the OECD principles each have a guide here. Because the EU's amending regulation of July 2026 changed parts of the AI Act's timetable after the Body of Knowledge was approved, read any legal date against the text it cites; our lessons say what changed.
Week 4, Domain III (8 hours). Development, in the Body of Knowledge's order: designing and building the system, including impact assessments and risk; data governance, lineage and provenance for training and testing; and III.C, release readiness, monitoring, audits and red teaming, incidents and public disclosures. The indicators here begin with "Govern", "Perform", "Conduct" and "Document", so practice the scenario questions, where a situation is described and you choose the next step.
Week 5, Domain IV (8 hours). Deployment: the context and the differences among model types and deployment options; assessing the system, including the vendor agreement; and IV.C, the largest competency on the exam, from deployment policies through monitoring and incidents to secondary uses, external communication and deactivation. Many of IV.C's indicators mirror III.C's for a deployed system, so study them together and notice what a deployer must do that a developer need not.
Week 6, the exam (about three hours beyond the 30). A timed mock in the exam's form first, 2.75 hours on the clock, then the domains the mock's section breakdown shows weakest, then the multi-select drill below. Book the exam before this week begins: appointments "must be scheduled at least 24 hours in advance", and a date on the calendar is the best study aid there is.
If you have more than 30 hours, add them to Domains III and IV first, then to II.C. If you have fewer, do not cut Domain I: it is 16 to 20 questions and the base of every scenario.
Practice the format, not only the content
Three features of the AIGP's format can cost a well-prepared candidate questions, and each can be rehearsed.
Multi-select questions. The FAQs explain that some questions ask for "a specific number of correct responses from the options provided (for example, 'Select 3 of the 5 options below')", that they are "clearly labeled on the exam", and that "No partial credit is awarded." Practice them as five separate yes-or-no judgments, one per option, and count your selections before moving on.
The two halves. The handbook says the break "divides the exam into two distinct halves, each containing half the total exam questions and half the total time limit", and "The candidate cannot return to the first half after submission." Rehearse the rhythm at least once: 50 questions on about 80 minutes, submit, a pause, 50 more. Our mock exam runs that way.
Scenario questions. The study guide says the questions "include case studies that present real-world challenges for AI model development and use." The Body of Knowledge says the exam sits "mostly" at the "remember/understand and apply/analyze levels" of Bloom's taxonomy, which means a scenario asks you to recognize which indicator applies and use it, not to design something new. When a stem is long, find the indicator first and the answer second.
What to study from
The IAPP's free materials cover the content. The FAQs list them: the candidate handbook, the Body of Knowledge and blueprint, a glossary of privacy terms, and "Key Terms for AI Governance", and they describe the paid materials, the textbook and training, as "tools to help you learn the information in the free resources."
The IAPP's training is a step, not a guarantee. The FAQs say so directly: "courses are not purely test prep, nor does taking training guarantee you'll pass your exam", and they explain why: its accreditation requires a separation between its training and certification departments. The handbook adds that the preparation materials "are not one-to-one guides of exam content."
Practice questions are where the format gets learned. The IAPP's practice exam is 100 questions as a PDF with explanations. Our AIGP practice questions live in the app, each with its rationale, the first competency free with a free account, and the course's question bank, flashcards and mock follow the Body of Knowledge competency by competency. What they are not is the exam: every question here is original, written in the exam's format. The handbook says any preparation materials from anyone other than the IAPP and its official training partners "are not endorsed by the IAPP", and that applies to this course. AIGP exam questions: what they look like shows the formats and explains why the handbook treats dump sites as cheating.
The week of the exam
The handbook's exam-day rules decide whether your 30 hours get used, so they belong in the plan.
- The window. Your exam "must be scheduled and taken within one year of purchase." Reschedule "up to 48 hours prior to the appointment time" in person, or "up to 15 minutes past the scheduled exam time" for OnVUE.
- In person. Arrive 15 minutes early with "two qualifying forms of ID". No devices and no reading material in the room; the timer does not stop for a restroom break.
- Online. A personal computer, no VPN, a system check of microphone, webcam and internet, and check-in that opens 30 minutes before the appointment with a photo of you and of a government-issued ID. "Candidates who are turned away for not having proper identification will forfeit all exam fees."
- The result. On screen at once, "Pass or Fail along with your score", then an email about the section breakdown by domain. A pass is 300 on the 100 to 500 scale, which the FAQs say "does not represent 60%". The AIGP exam: format, multi-select and the 300 explains the scale.
- If it goes wrong. A retake is a new purchase at a discounted price, no sooner than seven days later, and the section breakdown tells you where to put the next 30 hours.
Starting today
Read the 58 indicators once. Book a date six weeks out, or whatever your hours allow. Then begin Domain I with the lessons and practice questions for I.A, which are free in full with a free account, and let the blueprint's ranges set the pace from there.
By QualExam editorial. Written against the certifying body’s published documents, pinned by hash so a revised edition shows up as a change.