Key Takeaways:

✓ AI takes on the admin – as PMs, we use AI tools to automate meeting notes and summarise actions, so we can focus on your strategy.

✓ AI can’t spot the hidden risks – AI deals in numbers and static info; a PM can spot trends, patterns and issues before they become a real problem.

✓ Interpersonal relationships remain crucial – AI can’t have a nuanced conversation with a client; it can’t pick up on mood, or tone of voice, and it can’t read between the lines. All of these things need a human.

There’s a conversation happening right now that, if you’re a junior project manager, is hard to ignore. Scroll LinkedIn for five minutes, and you’ll find someone arguing that AI can now summarise meetings, draft status updates, build timelines and chase actions, so what’s left for a junior PM to do?

I’m a Junior Project Manager at The Digital Maze, and I’ve been thinking about this. I use AI every single day, but I don’t think it’s coming for my job, and I want to explain why, using what actually happens on my desk.

The Role of a Project Manager in a Digital Agency

As a Junior PM, my role is to ensure that our clients’ web projects proceed on time and to budget. This involves liaising with the client and our internal web design, development and digital marketing teams to make sure that all requirements and key actions are captured. I’ll also hold regular update meetings with the project teams to make sure everything stays on track, as well as managing any changes and revisions.

Curious about how project management differs from client management? Read our post on The Role of a Project Manager vs an Account Manager at a Digital Agency

What I Actually Use AI for as a Junior PM

Let’s start with the bit people skip past when they’re trying to make a point. I’m not anti-AI. I use it every day.

I use AI notetakers to transcribe and summarise client calls, so I’m not scribbling notes while trying to actually listen to what someone’s saying. I also use AI tools to draft a status update or an email when I know roughly what needs to go in it but want a clean starting structure.

Ultimately, I use AI to help me save time and work more efficiently, but never to do my job for me.

The Risk AI Wouldn’t Have Spotted

On some projects, we get more rounds of feedback back from the client than we plan for. Nothing dramatic on their own, just one more round here, a slightly bigger one there… On paper, the project is still on track – the timeline says so!

But, as a PM, I will notice the amount of feedback coming in doesn’t match the time we would actually schedule to review and act on it. AI would look at the same status updates and see a project on track, because that’s what the numbers say in that moment. It takes knowing how a project actually moves, not just what a tracker shows, to see that just two more rounds like that would ruin the timeline.

So I flag it early, reschedule the time we have to fit and let the client know straight away what that means for the date we’d originally given them. No surprises three weeks later and no scrambling to catch up – all avoided with an early and honest conversation.

That’s not a skill you can prompt your way into. It comes from actually watching enough projects to know when “on track” is about to stop being true.

The Work that Happens off the Page

A lot of project management doesn’t happen in the overall updates. It happens in a Slack message to a designer who’s gone a bit quiet, or when you notice that a developer’s running behind schedule because they’ve hit a hurdle. It happens in a one-on-one conversation where you ask someone how they’re getting on, and you can hear, before they even finish their first sentence (or even their greeting!), whether they need help or just need to vent.

AI can tell you a task is overdue, but what it can’t tell you is why someone’s stuck, or whether they need reassurance or more time, or if they just need someone to say, “Don’t worry, I’ll ask the client.” That’s a judgment call I make in real time, multiple times a day, based on the people I work with.

The Knowledge that Isn’t Written Down Anywhere

Every client has a personality: some want every decision explained in detail; some want the short version and trust you to have handled the rest; some go quiet when they’re unhappy rather than saying so directly, and you learn to read that silence for what it is.

None of that is written down in a brief or a project plan. It’s something that lives in me, built up over meetings, emails and the small moments and conversations that don’t make it into any report.

AI can read a transcript, but it can’t tell you that this particular client says “that’s fine” when they mean “I’m not thrilled, but I’ll let it go for now”, or that another client needs a video call instead of an email when something’s gone wrong, because that’s just how they prefer to get bad news.

So, Is AI Coming for Junior PM Roles?

I don’t think AI is coming for my role – I think it’s coming for the parts of the role that were never really the point; the transcribing, the first-draft writing, the small admin tasks that add up. Good riddance, honestly!

That’s time I can now spend on the parts that AI genuinely can’t do: reading a client, managing people and catching the thing that hasn’t gone wrong yet but that you know is just about to.

If you’re a Junior PM feeling anxious about where this leaves you, that’s completely natural!

But don’t worry. The value in your role has never been “I can produce documents.” It’s always been judgment, relationships and the ability to spot what’s really going on underneath what’s being said. AI hasn’t touched any of that, and I don’t think it’s about to any time soon.