The PMP now tests AI. Here is what that means for project managers

On July 9, 2026, the PMP exam started testing the use of AI in project management. That is PMI updating its own benchmark, and it settles a question project managers have circled for two years. AI skill is no longer optional in this profession. The exam that defines the job now assumes it.

What an exam change actually means

A certification exam is a job description with a pass mark. When PMI decides the PMP should test AI in project management, it is saying, in the most formal way the profession has, that a competent project manager works with AI. Hiring managers read exams that way. So do PMOs writing role descriptions and recruiters screening resumes. The PMP still certifies the breadth of the discipline: scope, schedule, risk, stakeholders, leadership. Working with AI is now part of that definition.

The gap the date exposes

The change tracks what organizations were already doing. PMI’s Pulse of the Profession finds 82 percent of senior leaders driving AI into project workflows, while about one in five project professionals reports solid practical AI skill. That is a wide gap between what leadership is pushing and what practitioners can show. Hiring reflects it. Microsoft’s Work Trend Index reports that 71 percent of leaders would rather hire a less experienced candidate with AI skills.

How to build the skill the exam assumes

Skip the survey courses. The skill in question is practical: can you hand a model part of your delivery work and stand behind what comes back? That only develops against real projects, so practice on yours.

• Rebuild one status report from your raw notes so a stakeholder reads it to the end, then mark every line you had to change
• Have a model sweep your live risk log for what is missing, then decide which of its additions are real and assign the owners yourself
• Ask for an estimate built from your own project history and check where it drifts from what actually happened
• Turn one set of stakeholder notes into a draft escalation and compare it against what was actually agreed

The correcting is the training. Each change teaches you where a model’s draft breaks on your kind of project. Write down what you changed and why. Keep a short list of the tasks you would never hand to a model unchecked, with a reason next to each. That record is worth more than any prompt, and it is exactly what an assessor looks for.

Where AI-PM fits next to the PMP

The PMP certifies the discipline. The Certified AI Project Manager (AI-PM) credential from the American School of AI certifies one specific practice: running status reports and risk logs with AI, and judging when the result is right. It sits alongside the PMP. It does not replace it.

AI-PM is a Stage 1 credential. The class is one day, live online, $495, and it is delivered by our authorized training providers. It runs five modules: where AI goes and where it must not; prompt patterns as a working library; weak versus strong, reading AI output like an editor; building your evaluation set; and the role brief. Of the day’s 390 minutes, 300 are hands-on on your own material.

The credential is earned after class. Within 14 days of the last day of class you submit a six-section pack built on your own sanitized work: cover block, prompt set, raw AI output, final version, change log and boundary note. For the AI-PM brief, that means one status report rebuilt so a stakeholder reads it to the end, plus a sweep of your own risk log listing what is missing and why. An assessor who did not teach the class, and who never sees your name or employer, reads it against a rubric published before purchase and returns a result inside 10 working days. The four criteria are Judgment, Specificity, Boundary awareness and Reusability, each banded Not yet, Pass or Strong. An unedited AI draft submitted as your own work lands in Not yet. The credential can be failed. One free resubmission with written feedback comes with the seat.

A pass is for life, with no renewal fee. You receive a printed Certificate ID and a digital badge, and anyone can check the record at Verify a Credential. AI-PM also admits you to Stage 2, Certified Advanced AI Coach, if you want to go further. The full path is on the certifications page.

If you hold a PMP, the July 9 change is a prompt to make your AI practice as visible as the rest of your resume. Read the AI-PM standard, compare its four criteria to your actual week, and start with the status-report drill above. It takes one Monday. When you are ready, find a class.

Sources: PMI (PMP exam change, July 9, 2026); PMI Pulse of the Profession; Microsoft Work Trend Index; American School of AI catalogue and certification standards

Frequently asked questions

Does the AI-PM credential replace the PMP?

No. The PMP certifies the breadth of the discipline: scope, schedule, risk, stakeholders, leadership. AI-PM certifies the specific practice of running that discipline with AI, assessed on work built from your own projects. It sits alongside the PMP.

How is the AI-PM credential assessed, and what does it cost?

The class is one day, live online, $495. Within 14 days of the last day of class you submit a six-section pack built on your own work. An assessor who did not teach the class reads it against a rubric published before purchase and returns a result inside 10 working days. A pass earns a lifetime credential from the American School of AI with a printed Certificate ID and a digital badge, and no renewal fee.

Can I fail, and what happens if I do?

Yes. Each of the four criteria is banded Not yet, Pass or Strong, and a Not yet on any of them means the credential is not issued. You get one free resubmission with written feedback. Pass rates for each stage are published after the first 30 marked submissions, and never before.