Project leadership in the AI era: things technology still doesn't do for us

Your team already uses AI in almost everything they deliver on the project. But is it actually helping people understand each other, or just making emails longer?

I'm running a project to bring processes and systems in line with a new legal requirement. It's a project involving a multidisciplinary team, with systems analysts, operations, legal, tax, sales, and representatives from practically every department in the company. And it's in this environment, with so many people in different roles working together, that I've been closely observing something worth sharing.

AI has fully entered the project's routine

AI has become part of every team member's daily work. It quickly understands what the legislation requires. It creates presentations and support materials to explain complex points. It scans the company's database looking for patterns and inconsistencies in records and helps fix that same database. In my role as PMO, I've been using it to put together status reports, summarize meeting minutes and organize information that used to take hours to consolidate.

Each of these workstreams has gained speed. Understanding that used to take days now happens in a few hours. Technical discussions move faster because everyone arrives with a similar level of information. So far, AI is doing exactly the job expected of it as a tool: speeding up technical work.

The point that stands out to me the most

There's one front, though, where that speed gain doesn't show up: the relationships between people on the team.

I notice this mainly in the emails exchanged between members of each area involved in the project. There's an easy-to-recognize pattern, long texts, full of details, explanations and justifications, with that characteristic structure of someone who used AI to write it. The result is that, on the other end, we also use AI, just to summarize the wall of text into two sentences.

And the cycle repeats. One person writes with AI, explaining and justifying, the other reads with AI, summarizing and simplifying. The exchange happens, but real understanding between the two people usually doesn't keep up.

The symptom shows up soon after. The same people who exchanged well-written, well-structured emails keep coming to the project manager to talk about what wasn't actually resolved. To understand how person A thinks, what person B really meant, why that reply didn't come back the way they expected.

I notice that direct conversation has been missing, getting up from the chair and walking over to the desk three meters away, making a call, pinging someone in chat to align, scheduling a video call to talk (when distance is a limiting factor). That's a move AI still doesn't make for the project.

Why project leadership remains irreplaceable in the AI era

This behavior isn't exclusive to the project I'm running right now. It tends to show up in any multidisciplinary team that's started using AI heavily in written communication, especially when the project involves areas with different interests, deadlines and vocabularies, as tends to happen in multidisciplinary initiatives.

That's exactly where the project manager's leadership gains an extra layer of importance. Not because AI gets in the way, but because it changes how people position themselves in writing, and that requires whoever leads to pay closer attention to what's behind each response.

Three things remain, in my view, the responsibility of whoever leads the project, and no AI does them for us:

• Having a feel for relationships: noticing when a well-written email hides a real disagreement between people from different areas, even when the tone sounds cordial.
• Understanding what's behind the message: separating what AI helped put into words from what the person actually thinks and needs, and using that to act before friction grows.
• Bringing the team together toward a shared goal: spotting dependencies between areas, facilitating the meeting that unblocks the conversation, and keeping everyone looking at the same target.

What this reflection leaves us with

Project leadership in the AI era is about the ability to use technology to speed up technical work without giving up on facilitating relationships, understanding conflicts, and keeping the team united around a shared goal.

AI speeds up a project's technical work in a way that would have been hard to imagine just a few years ago. That's good and should keep being explored by anyone managing projects. But it doesn't replace the role of whoever facilitates relationships, recognizes contributions, aligns a shared vision of the future, challenges how things are done, and celebrates wins together with the team.

These remain responsibilities the tool doesn't take on for us. That's exactly why a project manager's leadership remains, perhaps now more than ever, one of the most important skills for anyone running a project.

The observations in this article reflect my own personal perceptions from my work as a project manager, not those of any specific company.

If this topic resonates with you, it's worth pairing it with two articles that speak directly to it: on how leadership is built through example and values, in lead by example and by values, and on how trust sustains relationships within a project team, in trust and relationships in projects.

Take care! Leandro | PMP | PMO and Project Manager