AI Doesn't Change Project Management. It Changes What Great Project Managers Do.
How Project Managers Can Refine Their Value with AI Instead of Being Replaced
By Barry O’Reilly
September 10, 2026
What if AI gave project managers more time to do the work that only humans can do?
For most of my career, project management has been associated with coordination.
Build the plan. Track the milestones. Run the meetings. Chase the actions. Update the status report. Escalate the risks. Keep everyone aligned.
All of that work matters.
But ask any experienced project manager about the moments they are proudest of and they rarely talk about producing a perfect project plan.
They talk about the problem nobody knew how to solve. The stakeholder conflict they helped untangle. The risk they spotted before everyone else. The difficult decision they helped a team make. The moment a group of people stopped arguing about activity and finally aligned around an outcome. That has always been the highest-value part of project management.
The problem is that administration has increasingly crowded that capacity out, and this is where AI presents an extraordinary opportunity. Not to eliminate project managers. To give them more capacity to actually practice project management at its best.
Artificial Organizations: “What Percentage of Your Time Is Spent on Creative versus Administrative Work?”
The Ratio Is Changing
Many project managers spend enormous amounts of time on necessary but relatively low-leverage activities.
- Meeting notes.
- Action tracking.
- Status reports.
- Reformatting updates.
- Finding documents.
- Reconstructing what happened in the previous meeting.
- Preparing presentations.
- Sending reminders.
None of these activities suddenly becomes unnecessary because AI exists. But increasingly, you don’t have to be the person doing all of them.
That creates a very different question, “What will you do with the capacity you get back?”
That’s a question Pete Anevski, CEO of Progyny, asked his organization as it began experimenting with AI.
Pete had personally been maintaining a single Word document for his meetings with every direct report, manually tracking risks, actions, concerns and commitments. The initial experiment was simple: introduce an AI meeting assistant into his one-to-ones.
Conversations could be synthesized automatically. Actions, owners and dates were captured. Context became easier to retrieve. Decision cycles shortened.
But Pete’s important move wasn’t technological. It was human.
He told his organization they weren’t using AI to eliminate people. They were using it to elevate their human skills. And then he challenged them to think about what they could do with the capacity AI created.
That is a useful question for every project manager.
| More of yesterday's work | More of tomorrow's work |
|---|---|
| Capturing meeting notes | Listening deeply |
| Producing status updates | Identifying emerging risks |
| Tracking actions | Removing blockers |
| Reformatting information | Synthesizing competing signals |
| Updating plans | Testing whether the plan still makes sense |
| Chasing information | Asking better questions |
| Reporting activity | Improving decisions |
| Coordinating tasks | Solving problems |
The opportunity isn’t to squeeze even more activity into your calendar. It’s to change the ratio.
Less administrative work, more creative problem solving and better judgement.
More time addressing the difficult problems successful projects are actually built around.
Project Management Is Moving Up the Value Chain
For decades, we’ve often measured project managers by how well they managed plans, milestones and reporting.
AI can increasingly support much of that work. That means the scarce capability is no longer coordination alone. It’s judgment.
The profession is moving up the value chain.
| Traditional Project Manager | Emerging Project Manager |
|---|---|
| Manages tasks | Solves problems |
| Tracks progress | Improves decisions |
| Reports status | Shapes outcomes |
| Coordinates people | Enables collaboration |
| Measures activity | Measures impact |
| Escalates risk | Anticipates uncertainty |
| Documents learning | Accelerates learning |
| Delivers projects | Builds organizational capability |
You could describe the evolution another way:
Manage Tasks → Manage Decisions → Manage Learning
Project management tools remain valuable for managing the work. But the project manager becomes increasingly valuable for managing the problems surrounding the work.
- What are we learning?
- What assumptions have changed?
- Where is risk increasing?
- What decision are we avoiding?
- Where are stakeholders interpreting the same information differently?
- What outcome are we actually trying to achieve?
Those aren’t software configuration questions. They’re leadership questions.
So Why Do Most AI Initiatives Start with the Wrong Question?
There’s an understandable tendency inside organizations to begin AI transformation with, “What AI tool should we buy?”
Should we use Copilot? ChatGPT? Gemini? An enterprise agent? A new project management platform? But that’s starting at the end.
In Artificial Organizations, I introduce a simple model I call Traits → Tasks → Tools, or 3T.
Artificial Organizations: “The 3T Model”
The sequence matters.
1. Traits: How do you do your best work?
Are you at your best,
- Facilitating conversation?
- Listening?
- Visualizing complexity?
- Asking questions?
- Analyzing information?
- Building relationships?
- Spotting patterns?
- Creating structure from ambiguity?
These aren’t incidental preferences. They’re part of how you create value.
2. Tasks: Where should that capability be applied?
Now examine your work.
Which tasks genuinely need you? Which tasks drain time without requiring much judgment?
I use a simple rule: Automate the repeatable. Amplify the creative.
In the book I separate tasks such as judgment, coaching, design, strategic thinking and decisions from administrative work such as typing notes, writing follow-ups, reformatting updates and hunting for information.
For project managers, that distinction is especially powerful.
3. Tools: What technology supports that work?
Only now should you choose the AI.
If your problem is losing actions after meetings, choose technology that captures them.
If your problem is synthesizing hundreds of project inputs, find technology that helps identify patterns.
If your problem is pressure-testing a recommendation before an executive steering committee, use AI as a thinking partner.
The tool follows the work. The work follows the human.
That’s very different from buying licenses and then desperately searching for use cases.
From Information Management to Decision Management
This distinction becomes even clearer when you look at the work of Steve Elliott, founder and CEO of Dotwork. Steve sold his last project management company AgileCraft to Atlassian, now renamed JIRA Align.
Steve noticed a strange pattern after decades of working with executive teams. Companies invest millions in sophisticated project and portfolio management systems. Yet when a really difficult decision appears, executives frequently reach for a spreadsheet.
Why?
Not because spreadsheets are sophisticated, because leaders are trying to think.
The difficulty wasn’t lack of information. It was synthesis.
Signals were scattered across tools. Different functions had competing narratives. Historical context disappeared as priorities changed. Leaders spent huge amounts of cognitive effort reconstructing what was actually happening.
And that’s where decision latency enters the system.
Steve found that while the content of strategic decisions changes, the shape of the decision is surprisingly consistent:
- What’s happening now?
- What’s changed?
- Where are signals converging or diverging?
- What assumptions are we making?
- What happens if we’re wrong?
Those are fantastic questions for a project manager too.
Imagine walking into your next steering committee having already used AI to examine the project’s meetings, metrics, risks, customer feedback and previous decisions.
Instead of asking, “Can you summarize all this?”
Ask, “What decision is this information pointing toward?”
That single shift moves AI from information production toward judgment support.
And it moves the project manager from reporting the work toward helping the organization decide what to do about it.
Artificial Organizations: “The Architecture of an Artificial Organization”
Five Experiments You Can Start on Monday
You don’t need an AI transformation strategy to begin. Choose one project, one meeting or decision to start with. Then run one experiment.
1. Stop taking meeting notes manually.
Use an approved AI meeting assistant to capture the conversation. Stay present. Listen for disagreement, hesitation and uncertainty rather than trying to type everything said.
2. Turn the meeting into a decision asset.
Afterward ask AI: What decisions were made? What remains unresolved? What assumptions appeared? Who owns the next action? What risks changed?
3. Prepare for your next steering meeting differently.
Give AI your recent meeting notes, metrics, risk register and key updates. Ask: “What has materially changed since our previous review? Where should leadership focus its attention?”
4. Pressure-test one important recommendation.
Before you present it, ask AI to argue against your preferred option. Ask what evidence would disconfirm your position and what would need to be true for the recommendation to fail.
5. Measure your ratio.
For one week, estimate how much time you’re spending on administration versus creative problem solving and judgment. Then pick one repetitive activity to redesign.
The goal isn’t to become an AI expert. The goal is to become a better project manager.
Your Project Management Tool Should Manage the Tasks
Your project management platform is very good at storing tasks. Let it. Your job increasingly lies somewhere else.
Projects rarely fail because somebody forgot how to create another task in Jira, Microsoft Project, Asana or Monday. They struggle because people disagree about what matters.
Signals arrive too late. Risks remain hidden. Assumptions go untested. Decisions get escalated unnecessarily. Learning fails to travel. Teams become busy while outcomes drift. That is the work project managers can reclaim.
Machines are extraordinarily capable of processing information. Humans remain responsible for deciding what matters. Artificial Organizations argues that the constraint in AI-augmented work isn’t access to information anymore. It’s judgment under pressure.
That should be exciting for the project management profession.
Because judgment has always been at the heart of great project management.
The Evolution of Project Management Has Already Started
Great project managers aren’t remembered because they wrote the best status reports.
They’re remembered because they helped people solve difficult problems together.
They brought clarity when things were confusing.
They saw risks before other people did.
They created alignment when stakeholders disagreed.
They helped teams make difficult decisions.
They learned and adapted when the plan stopped matching reality.
AI creates an opportunity to spend less time documenting that work and more time doing it.
That’s why the transformation shouldn’t begin by asking which AI tools your project managers need.
Start by asking, “What do we want our project managers to become better at?”
Then examine the work preventing them from doing more of it.
Automate the repeatable. Amplify the creative.
Let your tools manage more of the tasks.
And give your project managers the capacity to manage the problems, decisions, risks, learning and outcomes that ultimately determine whether a project succeeds.
The future of project management isn’t about managing more projects.
It’s about helping organizations learn faster, make better decisions, and deliver outcomes that matter.
Barry O’Reilly
To learn more about Artificial Organizations: Build Better Judgment, Speed, and Results with Human and Machine Intelligence get a copy here.
To learn more about Barry O’Reilly, go to barryoreilly.com.
