An AIGP study plan built from the item weights, not from habit

40 hours of study should split roughly 8, 10, 11 and 11 across the four AIGP domains, in that order. That allocation comes from the item ranges in the IAPP's own exam blueprint, not from where the reading material is thickest. Most candidates do something close to the reverse, and the numbers say they are wrong.

What the blueprint actually weights

The AIGP Body of Knowledge and Exam Blueprint, version 2.1, was approved on 9 September 2025 and takes effect on 2 February 2026. It assigns each domain a range of items, not a fixed count. Take the midpoint of each range and you get a defensible estimate of where the exam's 100 items sit.

Domain	Item range	Midpoint	Share of midpoints	Of 40 hours
I. Foundations of AI governance	16 to 20	18	21%	8
II. Laws, standards and frameworks	19 to 23	21	25%	10
III. Governing AI development	21 to 25	23	27%	11
IV. Governing AI deployment and use	21 to 25	23	27%	11

The ranges sum to between 77 and 93, so they are not a literal item count. But as relative weights they are unambiguous. Domains III and IV carry 46 of the 85 midpoint items between them. That is 54 percent of the exam riding on how AI systems are actually built, released, assessed and used.

Where the habit comes from, and why it fails

The legal domain attracts study hours for a simple reason: laws are written down. You can read a statute, highlight it, summarize it and feel productive. Governance practice is harder to study from a text, so candidates drift back to Domain II and read it a third time.

Look at what that buys you. Domain II is worth 19 to 23 items. Even perfect performance on every one of them leaves at least 77 items you have not touched. Meanwhile Domain III alone matches or exceeds Domain II on every point of its range, and Domain IV does the same.

The competency ranges sharpen the picture further. The two heaviest competencies in the entire blueprint are both operational:

Competency	Item range
IV.C Govern the deployment and use of the AI system	9 to 11
III.C Govern the release, monitoring and maintenance of the AI system	8 to 10
II.C Understand the main elements of AI-specific laws	6 to 8

IV.C by itself can carry as many items as the whole of II.A and II.B combined, which cover how data privacy laws and other existing laws apply to AI at 4 to 6 items each. If your notes on post-deployment governance are thinner than your notes on privacy law, your effort is inverted relative to the exam.

The 1.8 minutes arithmetic

The IAPP describes the format plainly:

The AIGP is a three-hour exam composed of 100 items. All questions are multiple choice [...] There are no essay questions. Each correct answer is worth one point.

Three hours is 180 minutes. Divide by 100 items and you have 1.8 minutes per item, with no time banked for review unless you run faster than that. The IAPP does not publish what the items look like beyond this, so plan for the arithmetic, not for a guessed style. The practical consequence: you cannot afford to reconstruct answers from first principles at the desk. Material you studied has to be retrievable in under 2 minutes, which means your study should end in recall practice, not in a final read-through.

Every item is worth one point. An item on monitoring a deployed model pays exactly what an item on a privacy statute pays. The only rational move is to buy points where they are cheapest for you, and for most candidates the cheap points are in the domains they skipped.

The 40-hour plan, by competency

Allocate hours in proportion to the midpoint of each competency's range, rounded to half-hour blocks.

Competency	Item range	Hours
I.A What AI is and why it needs governance	4 to 6	2
I.B Organizational expectations for AI governance	5 to 7	3
I.C Policies and procedures across the AI life cycle	6 to 8	3
II.A Existing data privacy laws applied to AI	4 to 6	2.5
II.B Other existing laws applied to AI	4 to 6	2.5
II.C Main elements of AI-specific laws	6 to 8	3
II.D Industry standards and tools	3 to 5	2
III.A Designing and building the AI system	6 to 8	3.5
III.B Data collection and use in training and testing	6 to 8	3.5
III.C Release, monitoring and maintenance	8 to 10	4
IV.A Key factors and risks in the deployment decision	6 to 8	3
IV.B Key activities to assess the AI system	5 to 7	3
IV.C Governing deployment and use	9 to 11	5
Total		40

Two competencies get the most hours, and neither is legal. IV.C gets 5 hours because it is the largest single block on the exam. II.D gets 2 hours because 3 to 5 items is the smallest range in the blueprint, whatever the standards documents weigh on your shelf.

Running it in evenings

40 hours is 20 two-hour evenings: 4 weeks at 5 evenings a week, or 7 weeks at 3.

Sequence matters less than proportion, but a sensible order is Domain I first, because life cycle policies and organizational expectations underpin everything in Domains III and IV. Then run Domains III and IV back to back, since development governance flows into deployment governance. Leave Domain II until last. It is where most candidates' existing knowledge is strongest, so it needs the least warm-up, and studying it late keeps its details fresh.

Inside each block, spend the final 20 minutes on recall from a closed notebook rather than rereading. The 1.8-minute pace rewards retrieval, and retrieval is trained, not absorbed.

One version note. Blueprint 2.1 supersedes 2.0.1 and is effective from 2 February 2026. If you sit the exam near that date, confirm which version applies to your booking before you finalize the plan.

The exam will not ask you where you spent your evenings. It will draw 21 to 25 items each from development and deployment governance whether you prepared for them or not, and it will pay one point for each, the same as it pays for the statute you can already recite. Build the plan from the ranges, hold the proportions when the legal reading tempts you back, and let the blueprint, not habit, decide where the last 10 hours go.