AI is transforming how work gets done. And as coordination, compliance, and reporting functions become automated, many leaders (including me) are wondering:
What role, if any, is left for the PMO?
“Gartner was pretty clear. They think PMOs are dying.”
A client of mine had just attended a Gartner-sponsored conference. Over lamb sliders, she shared with me what she heard—in formal presentations, but also backstage. The message was blunt: as AI takes over coordination, compliance, and reporting, PMOs are on the way out.
I’d already read the press release they published in June, which predicted that 80% of today’s project management tasks would be eliminated by 2030. So I wasn’t shocked that was their position.
But then I reflected on a few months ago, when I heard that a publicly traded company I’ve worked with a few years ago announced they were disbanding their PMO. They laid off nearly all of their trained PMs and decentralized the rest.
As a person who works quite a bit with PMO leaders, I knew it was time to wrestle with a tough BDQ:
Are PMOs going away?
What Do We Mean by “PMO?”
Of course, there’s a question that has to come first. What, exactly, do we mean by PMO? Having worked with many of them, I can tell you there are about as many flavors as a Baskin-Robbins ice cream spread.
But let’s start with vanilla.
For the purposes of this discussion, I think we can make a few assumptions about what most people mean when they ask this question.
Here’s what I assume they’re picturing:
- A formal entity with a name, a structure, and a defined charter
- Staffed with two or more project managers
- Responsible for defining and enforcing how significant projects get done (through templates, processes, or policy)
- Tasked with reporting on project portfolio health and results to top leadership
Of course, PMOs can look very different from this. Some PMOs have no staff other than the person leading them. Some PMOs have only minimal governance functions. Some PMOs are totally external to a company (i.e., PMO-as-a-Service).
But formal, staffed, directive PMOs are the ones in the crosshairs of thought leaders and prognosticators.
So let’s talk about those.
Why Some Leaders Think PMOs Are Obsolete
First, I don’t think AI started the “PMO is dead” discourse. It just gave it new teeth.
For years, PMOs have faced side-eye from the organizations they were designed to help. And I’ll be really honest:
A lot of that criticism is fair.
Too many PMOs obsess over enforcing process, but can’t prove they deliver better outcomes than a business unit hiring its own PM.
If you want to see this dynamic up close, look at how PMOs approach change.
Too many PMO leaders train their PMs to be human shields against change, even in “Agile” projects. They want their PMs to execute a project “successfully” (i.e., on time), put a Win on the board, and move on to the next one.
Shifting priorities, new ideas, better clarity on requirements—as these inevitably unfold, PMOs perceive them as a threat to the baseline scope, schedule, and budget.
At best, this is a missed opportunity to be real business partners who ensure that projects deliver the intended benefit—inclusive of (but not limited to) adherence to baseline estimates.
At worst, this feeds into the narrative that the PMO is out of touch with what the business really needs—and that they’d be better off on their own.
Will AI Make Some PMOs Unnecessary?
When leaders hear that AI can write charters, monitor status, flag risks, and even summarize stakeholder sentiment…they start to wonder:
Do we really need a whole department to manage this anymore?
And here’s my honest answer. In many cases, you don’t.
Some organizations don’t need a Directive PMO. And with AI in the mix, the Directive PMOs, whose primary functions are project coordination, governance, and reporting, could probably scale way back.
But this doesn’t scare me, even as someone who often consults with PMOs. Here’s why.
1. Most organizations still need project management capability at scale.
Nearly every sizable organization needs a way to execute complex, cross-functional work reliably.
They need people who can:
- Translate strategic goals into executable plans
- Facilitate collaboration across silos
- Anticipate and plan for risk (especially risks that can’t be statistically quantified)
- Surface tradeoffs and enable smart decisions
- Elicit and manage requirements, detecting and responding when business needs or conditions on the ground shift
- Explain to stakeholders what is working, what isn’t, and what they can do to improve the odds of success
And while AI may assist with these tasks, it doesn’t replace the human instincts (and warmth) needed to navigate ambiguity, align and support stakeholders, and sense what is getting in the way of good execution.
So even if the acronym disappears—or gets replaced by something flashier like “Value Office” or “Execution Hub”—the capability is non-negotiable.
Companies will need to figure out for themselves—not from a Gartner think piece—whether a traditional PMO is still the most efficient way to do projects at scale.
And if it isn’t, they’ll need to figure out how to build enterprise project management capability without one. Which brings me to my next thought:
2. Some PMO models will decline. But others will thrive.
Not every organization needs a full-sized, directive PMO. But that doesn’t mean the whole concept is obsolete.
Here are a few PMO models I think will do just fine. Even as AI reshapes the landscape.
- Community-of-Practice PMOs. These don’t run the projects. They support the people who do. The PMs are embedded in business units, but the PMO provides shared tools, coaching, and structure. It’s light, flexible, and works well when centralized control isn’t needed (or wanted).
- One- or Two-Person PMOs. These folks don’t carry project load. Their job is to design the system: tools, playbooks, frameworks, escalation paths. If tools (including AI) are used, they help business units maximize the benefits, and mitigate the risks of the technology. They influence execution without being in the middle of everything—and they’re often the ones who keep things from falling through the cracks.
- Big Tent PMOs. Some PMOs are actually getting bigger—but smarter. I’m seeing more bring BAs, process improvers, testing leads, and change managers under one roof. The goal isn’t more control—it’s better coordination, faster throughput, and cleaner outcomes.
- PMO-as-a-Service Instead of staffing a permanent team, some organizations are moving to a “rent what you need” model—a kind of hybrid between a consulting firm and capacity partner. These groups provide PMs as needed, but they also recruit, vet, develop, and actively manage them. They share best practices across clients, offer just-in-time coaching, and sometimes even provide training for internal teams. It’s a solid option for companies that do a lot of project work but don’t want to build centralized PM capability in-house.
And full disclosure…I’m putting my money where my mouth is here. PMO-as-a-Service is something that The Persimmon Group offers, and I’m making a strategic bet that we’ll do more of it over the next five years, particularly for smaller and mid-size firms.
How PMO Leaders Should Respond to AI
If you lead a PMO—or influence one—stop doomscrolling. Here’s what I’m coaching my clients to do instead:
- Think in terms of strategy execution, not project management. Start with what your organization actually needs to move forward—not what the PMO “should” look like. Be willing to shift or reframe your role if that gets the work done better.
- Tailor your PMO to the organization—not the other way around. The best PMOs flex. They don’t push structure for structure’s sake. They listen, test, and adapt. Some of the best ideas for running projects at your organization may be happening outside your PMO, in the business. Listen for opportunities to transform ground-level innovation into enterprise practice.
- Build hybrid skill sets across your team. Encourage PMs to cross-train in business analysis, change management, or other “sister disciplines” that amplify their value (even if they don’t play multiple formal roles). Don’t think about your PMO as a team of PMs. Think of them as a team of cross-functional strategy execution experts.
- Hire fewer rule followers. Hire more rule definers. You want people who know how to build good, flexible processes—and who can tell when to adapt them. People who understand that standardization is a means, not the goal, and that processes should be evolving along with business needs.
- Be a translator, not just a tracker. Your value isn’t in collecting status updates. It’s in helping the org make sense of what’s happening and what needs to happen next. PMs should be doing this within their projects, but you should be doing it at the PMO/executive level.
One Question Every PMO Leader Should Ask
Look at what your PMO does today and ask:
If we stopped doing this tomorrow, who would notice—and why?
Whatever passes that test is the work worth evolving.