PMI Certified Professional in Managing AI credential artwork

PMI Certified Professional in Managing AI (PMI-CPMAI) Exam Guide

Validates knowledge of managing AI projects from business understanding and data preparation to deployment and governance.

PMI-CPMAI tests whether you can guide an AI project through business understanding, data, model development and deployment while keeping it responsible and compliant. Work through the required course, then practise scenario questions for each domain.

Understand the exam

Syllabus

Responsible AI appears across all five domains.

ISupport responsible and trustworthy AI efforts15%
  1. Task 1Oversee privacy and security plan
  2. Task 2Manage AI/ML transparency
  3. Task 3Conduct bias checks
  4. Task 4Monitor regulatory and policy compliance
  5. Task 5Manage accountability documentation and audit trail

Example scenario: A loan approval model performs worse for applicants from one region. What should you check, and who should be involved?

IIIdentify business needs and solutions26%
  1. Task 1Identify problem to be solved
  2. Task 2Evaluate initial AI feasibility
  3. Task 3Conduct risk assessment(s)
  4. Task 4Develop AI project scope statement
  5. Task 5Determine ROI
  6. Task 6Manage adoption/integration risks
  7. Task 7Draft AI solution
  8. Task 8Define success criteria
  9. Task 9Support business case creation
  10. Task 10Identify project resources

Example scenario: A manager wants a chatbot to cut support calls, but nobody has measured why customers call. What is your first step?

IIIIdentify data needs26%
  1. Task 1Define required data
  2. Task 2Identify data SMEs
  3. Task 3Identify data sources and locations
  4. Task 4Coordinate AI workspace and infrastructure
  5. Task 5Gather required data
  6. Task 6Check data privacy, compliance, and access
  7. Task 7Oversee data evaluation
  8. Task 8Determine if data meets solution needs
  9. Task 9Convey data understanding to leadership

Example scenario: The data you need sits in three systems with different customer IDs and a year of missing records. What do you report to the sponsor?

IVManage AI model development and evaluation16%
  1. Task 1Oversee AI/ML model technique(s)
  2. Task 2Oversee AI/ML model QA/QC
  3. Task 3Manage AI/ML model training
  4. Task 4Manage data transformation to conduct data preparation
  5. Task 5Verify data quality for go/no-go decision to conduct data preparation
  6. Task 6Verify model ready for operationalization go/no-go decision

Example scenario: A model scores well in testing but the business team cannot explain its errors. Is it ready to deploy?

VOperationalize AI solution17%
  1. Task 1Manage creation of AI solution deployment plan
  2. Task 2Manage AI solution deployment
  3. Task 3Oversee model governance
  4. Task 4Oversee AI solution metrics
  5. Task 5Prepare final report/lessons learned
  6. Task 6Manage AI solution transition plan
  7. Task 7Oversee AI solution contingency plan

Example scenario: Three months after launch, the model’s accuracy is drifting downward. Whose job is it to notice, and what happens next?

PMI-CPMAI Exam Content Outline · September 2025

Format

Questions
120
Duration
2h 40m
Delivery
Test centre or online
Languages
English, Arabic, Chinese (Simplified), Chinese (Traditional), French, German, Japanese, Korean, Portuguese, Spanish

Question types

Scenario-based questions. Each question describes a situation on an AI project and asks for the best next step or decision.

CertUp example

A team has a promising churn model, but the business has not agreed how success will be measured. What should the project lead do before more model work?

  1. Tune the model for higher accuracy
  2. Agree the business success criteria and how they will be measured
  3. Deploy the model to a pilot group
  4. Collect more training data
Show answer

B. Without agreed success criteria the team cannot judge whether the model solves the business problem, so this comes before further iteration.

Eligibility and booking

Required before the exam

PMI-CPMAI Exam Prep Course

Required

Complete PMI’s 21-hour self-paced course. It is sold with the exam, and the exam unlocks when you finish.

Buy the course and exam bundle in your myPMI account. Schedule the exam with Pearson VUE once you finish the course.

Make a study plan

The required course is most of the work. Review the outline afterwards, then use the official practice exam to find weak domains.

Most candidates spend 30–60 hours. Plan for the low end if you already run data or AI projects; add time if AI project methods are new to you. Includes the 21-hour required course.

4 weeksto be ready
  1. 1Understand the examRead the syllabus and format, check you’re eligible and pick a target date.2 h
  2. 3Learn the core conceptsWork through one main course or book, topic by topic.24 h
  3. 4Practice with mock examsTake timed practice tests and review every mistake.12 h
  4. 5Prepare for exam dayBook the exam and check the ID and delivery rules.2 h

Learn the core concepts

The required PMI course is your main study resource. Use the outline to check you have covered every task.

See all 2 PMI-CPMAI courses →

Practice with mock exams

PMI’s practice exam is the only official question set. Retake it after reviewing the explanations.

See all 4 PMI-CPMAI practice tests →

Are you ready?

  • You can explain what happens in each CPMAI phase and who decides at each go/no-go point.
  • You score well on the official practice exam after reviewing your mistakes.
  • You can concentrate for 160 minutes without a break.

Prepare for exam day

Schedule with Pearson VUE once you finish the course. PMI recommends a test centre; online proctoring is also available.

Before travelling to the test centre

  • Bring a current government-issued photo ID with a name that matches your PMI account.
  • There are no scheduled breaks in the 160 minutes.
  • The tutorial and survey before and after the exam do not count toward your time.
  • Official results appear in your myPMI account. You can retake up to three times within a year, paying for each retake.
Pearson VUE for PMI

PMI-CPMAI Exam Content Outline

Advice from exam takers

Use the official practice exam as your benchmark.

The same candidate scored 85% on the first attempt and 95% on repeats, then passed, and found it closer to the real exam than third-party question sets.

Some find the course heavy on tool names.

One widely discussed post criticized the required course for focusing on AI products and tools rather than project process.