The adoption question has been answered. What's next?
Alex Rodov · PMO Leadership, PMP, Microsoft MVP Alumni
Not long ago, "Are you using AI?" was a real question in PMO circles.
Today, it sounds a bit like asking whether a team uses email.
The numbers tell a similar story. Tempo's 2026 State of AI in Portfolio Management report, based on responses from more than 300 PMO and project leaders, found that 91% were already piloting or using AI. A DesignRush roundup of project management conferences reported adoption figures rising from 36% to 70%.
The surveys measure different things, so I wouldn't put too much weight on comparing the numbers directly. But the direction is clear.
AI adoption is no longer the interesting question.
What matters now is whether AI is actually changing how we work.
Using AI and Getting Results From It Are Different Things
The more interesting finding in the Tempo research isn't the 91%. It's what comes after it.
Some teams are seeing better results than others. The difference comes down to the practices they've adopted and the specific tasks they've given AI, not simply whether they have access to it.
That's a real shift.
The gap used to be between teams that had AI and teams that didn't.
Now it's between teams that gave AI a specific job and teams that handed out logins and hoped for the best.
That's my read, anyway.
The question is no longer whether your team uses AI. It's which job you gave it.
For project managers, that distinction matters.
Giving someone access to an AI assistant doesn't automatically improve resource allocation, shorten reporting cycles, or identify project risks earlier.
Those improvements happen when teams identify a real problem, apply the right capabilities, and measure whether the work actually gets better.
AI adoption is a starting point. It isn't a result.
A Standard Has Arrived
PMI has published The Standard for Artificial Intelligence in Portfolio, Program, and Project Management, which it describes as the first global standard for AI in our profession.
PMI's webinars this summer emphasized a point I think sits at the heart of all this: AI initiatives can stall when teams focus on the technology and underestimate governance, data readiness, stakeholder alignment, risk, change adoption, and value realization.
Read that list again.
Almost none of it has to do with the model itself.
It's the work PMOs have always needed to do.
That's genuinely encouraging. The profession doesn't need to learn an entirely new craft so much as point an established one at a new tool.
Project managers already know how to define outcomes, manage risk, coordinate stakeholders, establish controls, and measure performance.
The challenge is applying those disciplines to AI-assisted work.
That means asking practical questions before rolling out another tool:
- What business problem are we trying to solve?
- Which task should AI handle or support?
- What data does it need, and is that data reliable?
- Who reviews its output?
- How will we measure whether it creates value?
Without those answers, AI risks becoming another item on the technology inventory rather than a meaningful improvement to project delivery.
The Thursday-Night Test
Here's a simple way to check whether AI has changed anything for your team.
Look at what happens the night before the weekly report.
A session at the University of Maryland's PM Symposium this year described project leaders spending more than ten hours a week searching for information across five tools and three spreadsheets, sometimes discovering risks only after the damage was done.
That's a presenter's description rather than a formal study, but I suspect it will sound familiar to plenty of project managers.
The data lives in one system. The schedule lives in another. Risk updates arrive by email. Financial information sits in a spreadsheet. Someone has to pull everything together before leadership can see what's happening.
Now imagine your team has adopted AI.
If Thursday night still involves the same manual collection, reconciliation, and formatting, the tool is in the building, but the work hasn't moved yet.
That isn't necessarily a failure.
It's a sign that adoption was only the first step.
The opportunity is to connect the information, automate the repetitive work, and give project managers a clearer picture of project health before the reporting deadline arrives.
The goal isn't simply to produce the same report faster.
It's to give people more time to understand what the report is telling them and decide what to do about it.
Three Questions for Your Team This Week
You don't need another AI strategy document to get started. Begin with three questions.
1. Which specific task did AI take off someone's plate?
Be precise.
Did it reduce the time needed to prepare a status report? Help identify dependencies? Summarize project risks? Improve resource forecasts?
If the person who used to do the work can't name what changed, the answer is probably none.
Start with one task. Establish a baseline. Measure the difference.
2. Who reviews an AI-assisted number before it leaves the team?
A person, with a name. Not a process nobody follows.
AI can produce useful analysis, but outputs still need appropriate validation, especially when they affect budgets, schedules, resource decisions, or contractual commitments.
Define who checks the output, what they verify, and when the result needs to be escalated.
Automation should reduce manual effort without removing accountability.
3. Could someone outside the project trust the data underneath it?
AI works with whatever it's given.
If project information is incomplete, inconsistent, or outdated, the resulting analysis may be unreliable too.
Before expecting AI to deliver better decisions, make sure the underlying data is fit for the job.
Clean inputs matter more than clever prompts.
The Next Step Is Turning Adoption Into Value
This is a good week to ask these questions.
Project Controls Expo USA wrapped up on Wednesday, with sessions covering earned value, schedule intelligence, and AI-assisted estimating. The PMI Global Summit opens in Detroit on October 21.
The conversations happening in those rooms circle the same point: the tools are here, and the craft is in how we use them.
For PMOs, the next phase of AI adoption isn't about collecting more tools or encouraging everyone to experiment indefinitely.
It's about connecting AI capabilities to specific business outcomes.
Less time spent assembling reports.
Earlier visibility into project risks.
Better-informed resource decisions.
More reliable forecasts.
And more time for project managers to do the work that actually requires their experience and judgment.
Those are the kinds of results worth measuring.
Tomorrow is Friday.
If the report that goes out took less of your team's week than it used to, AI is doing its job.
If it didn't, you now have your first question for Monday.
Adoption was the easy part. Choosing the job is the craft.
Alex Rodov
Top comments (0)