Originally posted on LinkedIn on January 26, 2026. Read and engage with the original post here.
Every PM I talk to eventually brings up the same inevitable question: ‘Is AI going to replace me?’
I totally get it. AI is getting eerily good at things that we always assumed were uniquely human. It writes code, analyzes data, generates schedules and drafts status reports that sound like they came from an actual person. It’s so convincing that I find myself getting worried that I might “disappoint” my AI chat agents when I don’t get something done correctly. And honestly, if you are a PM watching this all play out, it’s natural to wonder where you fit in.
I’ve been taking a planned career break over the past 6 months. This has given me the chance to rest, recuperate, spend time with my family, and pursue some passion projects (like, uh, playing golf). I am learning and leveraging AI as I work towards building my own business. This is important to me because I am living the AI transformation, not just reading the theory behind it. I am convinced of the following:
AI will not replace project managers. But it will fundamentally change how we spend our day. And honestly? That’s a good thing.
If we aren’t filling our days with status report compilation, meeting note synthesis and endless tracking, what is left is what really matters. I wrote recently about how PM work has always been about the people. The human work. This is work no algorithm can match.
Here is how I see the shift playing out in practice.
The Grunt Work is Going Away
Not in 5 years. Not 18 months from now. Right now. Today. I have been using both Claude and ChatGPT recently, and they are both solid at cranking out traditional PM work that used to take up most of my time.
Status Reporting
I used to spend the first couple of hours of my day compiling updates across five different sources (spreadsheets, e-mails, chats, posts, etc.) into coherent status reporting for myself, the team, and stakeholders. Now, I feed meeting notes and status messages directly into Claude and it generates a draft in 30 seconds. I spent 10-15 minutes refining the messaging, clarifying, and looking for inconsistencies, but this aspect of project management can be significantly streamlined.
Getting an 80% solution quickly and then iterating is a game changer.
Meeting Note Summarization
After a 90-minute stakeholder meeting, I’d spend another 30 minutes writing up the action items and notes. I would take notes in the meeting, but they were never ready to share out and were often rough because I was trying to be an active participant in the meeting itself while taking notes. AI can generate concise, easy to read notes and clear action items in seconds. It often catches action items I missed because I was focused on facilitating, not documenting. It notices when someone says ‘I’ll look into that’ – things that slip by in the moment but matter later.
Risk Pattern Recognition
This may be my favorite current capability of AI in aiding project management. The ability to spot risk patterns in a project or even across programs and identify common themes that can be mitigated.
This can be quite subtle. For instance, AI can detect slight changes in tone/language coming from the team’s written messaging which may indicate a hidden risk. For example, if the status updates from the engineering leads shifts from highly technical to “making good progress” there may be a hidden bottleneck not being surfaced.
AI risk pattern recognition can spot things WE are doing/saying as PMs that may indicate underlying issues were aren’t seeing (or don’t want to recognize). For instance, it may detect we are using “hedging” language patterns like “we should be able to meet the launch deadline.” Getting this real-time feedback as early as possible on a project is super helpful.
Schedule Optimization
When it all comes down to it, optimizing a schedule, tracking buffer time and managing critical path are all math problems. AI can map dependencies and do a better job than just using spreadsheets.
Again, this isn’t to suggest that this is all automated and PMs aren’t having to do any work with schedules. But getting to 80% rough solutions quickly is much faster with AI than doing so manually, which gets us to the real work of iterating on fleshed out plans much quicker.
Everything I mention above is real and at your disposal today. If you are not using these tools, you are losing precious time you could be spending in areas of greater impact. If your value as a PM primarily comes from these types of tasks, it’s time to evolve.
But here’s the thing. These were never the most important parts of the job anyway.
The Irreplaceable Human Work
I can remember fondly many times when I walked into a tense project meeting. Emotions are high, stakes are immense, and opinions are running rampant. And it’s pretty clear this is not the environment where Claude or ChatGPT are going to help much. Here is what AI currently CAN’T do.
Read The Room
You are in a meeting and you can just sense something is off. The engineering lead and product manager both *say* they agree on the best path forward, but their body language is just screaming conflict. Not everybody in the room notices, but you do. AI is not going to be of much help in this situation.
I remember a situation where we had a decision-making meeting with the aim of moving forward quickly. I *thought* we were all aligned and had a plan and then I noticed that one of the top performers on the team was unusually quiet and had his arms folded while looking out the window. After the meeting, I asked if they were on board with the plan and he said that he didn’t feel like he was being heard in the meeting and so he shut down. We grabbed a coffee and I asked him for his POV on the best next steps. As we discussed the pros and cons of the solution he had in mind and the one the team had landed on, he realized that there were some drawbacks to his plan. We then discussed how we could work together to make sure all voices in the room were heard. From that point on, this team member was rarely (if ever) quiet during meetings and proactively shared his opinions.
AI ain’t doin that!
Navigating Politics
Let’s face it: we’re all human. And as organizations scale, political dynamics inevitably show up. There are invisible power dynamics that don’t show up in any org chart or project plan, and yet they can be *felt* by the people involved. Who *really* makes big decisions? Whose opinion actually moves the room? Who is the person that walks into a conference room full of lively conversation, only to have the group fall silent the moment they walk into the room?
AI can’t navigate any of this. It requires institutional knowledge, pattern recognition, and trust.
Making Judgement Calls
It would be great if you could develop an algorithm to manage complex projects, but what do you do when you have conflicting data? A customer demands a solution in 24 hours that your engineering team says will take 3 weeks. What do you do? AI can analyze and present the options, but it can’t make the call for you. Proper judgement calls require knowing the context of the situation, intimate knowledge of the people involved, the risk tolerance of the company and/or customer and what is actually at stake.
Sometimes the right call is to disappoint the customer in the short term to preserve the relationship long term. Or to push back on engineering to find a creative compromise. These judgment calls require wisdom AI doesn’t have.
AI didn’t break bread with the customer and get to know them as people, delving deeply into their actual needs. You did. This puts you in the best position to make judgement calls.
What This Means For Us
So if AI can handle tracking and reporting, what is left for us? We are shifting from ‘person who knows all the details’ to ‘person who sense what’s really happening.’ There is a massive delta between tracking what is happening and *interpreting* what is occurring. The project manager of the future will be an organizational therapist who makes sense of the chaos in ways no algorithm can.
Traditionally, one of the most important jobs of project managers was the definition and championing of the process. We were the ones that made sure the sprint meetings took place as planned, we held the team accountable to provide status reports on time. AI will be able to run ceremonies, but it can’t create an environment where people surface problems early and transparently. It can’t build the relationships that make people want to help each other. Only a human organizational therapist can do that. This is the special sauce, the differentiator that the human project manager will bring.
In the world we are shifting to, we must prioritize empathy over efficiency, sensing over tracking, judgement over analysis, and trust-building over process adherence. We must embrace AI as a co-pilot, not a competitor. Leverage the tooling to save time and then reinvest that extra bandwidth into what matters most. The people.
And honestly? That’s the job I want. That’s the job that matters.
So when people ask me ‘Will AI take your job?’ I tell them: No. But it will give me back the time to do my job properly. To focus on what actually moves projects forward – people, not process.
The PMs who figure this out early will have an advantage. The ones who cling to the old way of working will struggle.
Which one will you be?”
What’s one PM task you’d love AI to take off your plate so you could focus on the human work? Drop a comment – I’m curious what resonates.”