- What You Are Actually Sitting: The 2026 Exam in Plain Terms
- Before You Can Schedule: The Course Gate and Eligibility Clock
- Reading the Blueprint: Where the Points Live
- Domain-by-Domain: What You Must Be Able to Do
- Do Not Confuse Methodology Phases with Exam Domains
- A Domain-Weighted Study Sequence
- Exam-Day Mechanics: Test Center vs. OnVUE
- Attempts, Retakes, and the 12-Month Window
- After You Pass: Careers and Renewal
- Frequently Asked Questions
- The current exam is PMI-CPMAI: 120 questions (100 scored, 20 unscored), 160 minutes, no scheduled breaks.
- Business Needs and Solutions plus Data Needs are 26% each, so 52% of the blueprint sits in two domains.
- The 21-hour prep course must be completed before you can schedule the exam.
- You have 12 months from purchase and up to three total attempts, not three retakes.
What You Are Actually Sitting: The 2026 Exam in Plain Terms
Cognitive Project Management for AI began life as a methodology from Cognilytica, which is now part of the Project Management Institute. On September 30, 2025, PMI introduced the PMI Certified Professional in Managing AI (PMI-CPMAI) to replace CPMAI v7. If you studied from older v7 materials, or you hear colleagues talk about the "+E" variant, treat those rules as legacy. They are not interchangeable with the exam you will sit in 2026. For a grounding in the name and history, see our explainers on what CPMAI is and what CPMAI stands for.
The format is straightforward to state and easy to underestimate:
| Element | What to Expect |
|---|---|
| Total questions | 120 (100 scored, 20 unscored pretest items) |
| Testing time | 160 minutes |
| Scheduled breaks | None scheduled on this exam |
| Delivery | Pearson VUE test center or OnVUE remote proctoring (subject to regional availability) |
| Book policy | Closed-book; built-in calculator in the computer-based testing environment |
| Result | Pass/fail with domain-level diagnostic feedback |
Two details matter for pacing. First, you cannot tell which 20 items are unscored, so every question deserves a full read. Second, the tutorial and survey screens fall outside the 160-minute timer, but there is no scheduled break inside it. Plan your hydration, caffeine, and bathroom habits accordingly. The exam-specific no-break rule overrides what you may have read in generic handbook descriptions for other PMI exams.
Before You Can Schedule: The Course Gate and Eligibility Clock
PMI sells the preparation course and the certification exam together as a bundle rather than as a standalone exam fee. You must complete the 21-hour PMI-CPMAI Exam Prep Course before scheduling and taking the exam. Note that 21 hours is the length of the course, not the length of the exam. Our CPMAI requirements guide covers eligibility in more depth, but the key points are:
- No prior credential or experience is required. You do not need a project-management background, a technical degree, or previous AI certification.
- Membership is optional. Pricing varies by region and membership status.
- The eligibility clock starts at purchase. You have 12 months from purchase to obtain certification.
On cost: the issuer pages we checked displayed loading placeholders where prices should be, so we are deliberately not quoting a dollar figure here. Confirm current bundle pricing, retake fees, and renewal fees directly on PMI's site, then compare against the framework in our CPMAI certification cost breakdown.
Reading the Blueprint: Where the Points Live
The exam content outline defines five weighted domains. These are the only weights that determine how the exam is built:
| Domain | Weight |
|---|---|
| Domain 1: Support Responsible and Trustworthy AI Efforts | 15% |
| Domain 2: Identify Business Needs and Solutions | 26% |
| Domain 3: Identify Data Needs | 26% |
| Domain 4: Manage AI Model Development and Evaluation | 16% |
| Domain 5: Operationalize AI Solution | 17% |
Exact question counts per domain can vary by form, so use the percentages as a guide to emphasis rather than a promise of a specific number of items. The headline is that Domains 2 and 3 together make up 52% of the blueprint. A candidate who is excellent at model evaluation but vague on data readiness is studying the wrong half of the exam. For a longer treatment, read our complete guide to all five CPMAI content areas.
Key Takeaway
Allocate your study hours roughly in proportion to the weights, then add extra time to whichever domain feels weakest after your first practice round. Because PMI reports domain-level diagnostic feedback, a weak showing in one area is visible, but the pass/fail decision rests on the overall standard.
Domain-by-Domain: What You Must Be Able to Do
The outline describes tasks under each domain and notes that its task enablers are illustrative rather than exhaustive. That means you should expect scenario questions that apply a concept rather than recite a definition. Below is how to think about each domain.
Domain 2: Identify Business Needs and Solutions (26%)
This is the "is AI even the right answer?" domain. Questions test whether you can connect a business problem to an appropriate AI approach and judge feasibility before anyone writes code.
- Translating a vague business ask into a defined problem and measurable success criteria
- Deciding whether an AI solution, a simpler analytics approach, or no technology change fits the need
- Assessing value, cost, risk, and organizational readiness for an AI initiative
- Recognizing which AI pattern (for example prediction, classification, conversational, or recognition-style use cases) matches the stated goal
- Engaging stakeholders and setting expectations about what AI can and cannot deliver
Domain 3: Identify Data Needs (26%)
The CPMAI mindset treats data as the first-class citizen of AI work. Expect scenarios where the correct answer is to pause and fix a data problem rather than push ahead with modeling.
- Determining what data an AI solution requires and where it will come from
- Evaluating data quality, quantity, relevance, bias, and labeling needs
- Understanding data access, privacy, governance, and security constraints
- Planning data preparation and pipelines, and recognizing when data is not ready
- Identifying data gaps that threaten feasibility early rather than late
Domain 4: Manage AI Model Development and Evaluation (16%)
You are a manager here, not a data scientist, but you must speak the language well enough to steer the work and judge results.
- Overseeing iterative model development and experimentation
- Selecting appropriate evaluation approaches and interpreting results against business success criteria
- Recognizing overfitting, underperformance, and when a model is "good enough" for the business case
- Deciding when to iterate, return to earlier steps, or stop
Domain 5: Operationalize AI Solution (17%)
A model that never reaches production delivers no value. This domain covers deployment, monitoring, and sustaining a solution after launch.
- Planning deployment, integration, and the transition into operations
- Monitoring performance and handling model drift and degradation over time
- Establishing feedback loops, maintenance, and ownership for the live solution
- Managing change for the people and processes that use the AI output
Domain 1: Support Responsible and Trustworthy AI Efforts (15%)
The smallest domain by weight, but its themes run through scenarios in every other domain, so do not treat it as optional.
- Applying principles of fairness, transparency, accountability, privacy, and safety
- Identifying ethical, legal, and regulatory risks in an AI initiative
- Supporting governance, documentation, and human oversight of AI systems
- Recognizing bias and its sources, and raising it at the right moment
Do Not Confuse Methodology Phases with Exam Domains
This is the single most common source of confusion for CPMAI candidates. The CPMAI methodology is organized into six iterative phases, and the prep course is organized into seven modules. Neither of those is the exam blueprint. The five weighted domains in the content outline are what the exam is built from.
- Prep course modules (seven) teach the material and are a curriculum, not a scoring structure.
- Methodology phases (six) describe the iterative lifecycle you will use to run an AI project.
- Exam domains (five) determine how many questions you see on each topic area.
Practically, this means you should finish the course, then re-map what you learned onto the five domains. A phase like data understanding contributes content to Domain 3, but a single domain may draw on several phases. Build your notes by domain, not by course module. Our CPMAI cheat sheet is organized the same way for quick review.
A Domain-Weighted Study Sequence
Here is the one structured plan in this guide. It assumes you have completed the 21-hour course and want to consolidate over roughly five weeks, sequencing domains by weight and by dependency rather than by generic habit. Adjust the pace to your own calendar and the 12-month eligibility window.
Domain 2: Business Needs and Solutions
- Rebuild notes on problem framing, feasibility, and AI pattern selection
- Practice scenarios where the right answer is "reframe the problem" or "don't use AI"
Domain 3: Data Needs
- Cover data quality, bias, labeling, access, and readiness judgments
- Drill the "pause and fix the data" scenario type
Domains 4 and 5: Development, Evaluation, and Operations
- Link evaluation results back to business success criteria
- Review deployment, monitoring, drift, and ownership after launch
Domain 1: Responsible and Trustworthy AI
- Study it after the others so you can layer ethics and governance onto concrete project scenarios
- Take a first full-length timed practice exam
Gap Closure and Simulation
- Return to your weakest domain based on practice results
- Do at least one 160-minute sitting with no break to build stamina
The reasoning: Domains 2 and 3 come first because they carry 52% of the weight and because later domains assume you understand problem framing and data readiness. Domain 1 comes later not because it matters less, but because responsible-AI judgments are easier to apply once you have a full project picture. Use our CPMAI practice tests to check readiness at the end of weeks 3 and 5.
Exam-Day Mechanics: Test Center vs. OnVUE
Pearson VUE delivers the exam either at participating computer-based testing centers or through OnVUE remote proctoring, subject to regional availability. The two experiences differ in ways worth planning for:
| Factor | Test Center | OnVUE Remote |
|---|---|---|
| Environment | Provided by the center | Your own private, compliant space |
| Preparation | Arrive with valid ID | System check and identity verification beforehand |
| Notes and writing | Follow center and administrator rules | Books, notes, paper, and physical writing materials prohibited unless specifically approved; a digital whiteboard is provided |
| Calculator | Built-in calculator in the computer-based testing policy | Built-in calculator in the testing software |
Do not assume what is allowed at a test center carries over to remote testing. Remote writing restrictions are stricter, and scratch materials supplied at a center are an administrator-provided exception, not something you can improvise at home. Run the OnVUE system check well before exam day, and rehearse using the digital whiteboard if you plan to test remotely. Because there are no scheduled breaks, confirm your setup is comfortable for the full 160 minutes. For scheduling windows and deadlines, see our CPMAI exam dates guide.
Attempts, Retakes, and the 12-Month Window
The attempt rules are frequently misquoted, so read them carefully:
- You have up to three total exam attempts within your eligibility year. That is three attempts in all, not three retakes after a first try.
- Additional attempts require payment.
- The outline recommends 30 days of preparation before a retake. This is advice, not a mandatory 30-day waiting period.
- After three failed attempts within the eligibility year, the general handbook requires a one-year wait from your last exam before you can reapply.
On scoring, PMI uses a criteria-based, psychometrically established passing standard and reports pass/fail with domain-level diagnostic feedback. No numerical passing percentage was verified, and no official pass rate was confirmed, so be skeptical of any site quoting exact figures. Our pages on the CPMAI passing score and CPMAI pass rate explain what can and cannot be said responsibly. If you do not pass, use the domain-level feedback to direct your retake preparation rather than rereading everything.
After You Pass: Careers and Renewal
PMI-CPMAI is aimed at professionals who lead or govern AI initiatives: project and program managers, product owners, business analysts, data and analytics leads, and consultants who translate between business teams and technical teams. Because the credential is non-technical in its entry requirements, it also attracts career changers moving toward AI delivery. We cover the market in CPMAI jobs, and the financial picture in our salary guide and ROI analysis. We do not quote salary numbers here because none can be verified from the issuer.
Maintaining the credential follows PMI's continuing-certification model. Renewal requires 30 PDUs every three years, structured as follows:
- At least 18 Education PDUs
- At most 12 Giving Back PDUs
- At least four PDUs in each Talent Triangle skill area: Ways of Working, Power Skills, and Business Acumen
- Report your PDUs and pay the applicable renewal fee
Since the minimums per skill area add up to 12 of the 30, you have flexibility in the remaining PDUs, but you cannot ignore any one area.
Frequently Asked Questions
No. There is no prior project-management, technical, or AI experience or certification requirement. You must, however, complete the 21-hour PMI-CPMAI Exam Prep Course before scheduling the exam.
No. Twenty-one hours is the duration of the required prep course. The exam itself is 120 questions (100 scored, 20 unscored) with 160 minutes of testing time and no scheduled breaks.
Identify Business Needs and Solutions and Identify Data Needs are weighted 26% each, together 52% of the blueprint. Start there, then cover Operationalize AI Solution (17%), Manage AI Model Development and Evaluation (16%), and Support Responsible and Trustworthy AI Efforts (15%).
You get up to three total attempts within the 12 months from purchase. Additional attempts require payment, and PMI recommends 30 days of preparation before a retake, though that is not a mandatory waiting period. After three failed attempts in the eligibility year, a one-year wait applies before reapplying.
Not unless specifically approved. OnVUE prohibits books, notes, paper, and physical writing materials, and provides a digital whiteboard instead. A built-in calculator is available under the computer-based testing policy, and you must complete a system check and identity verification beforehand.