Ben Webb on AI in Project Management: Why the Hype Is Ahead of the Reality

 

A lot has been made of AI in project management. 

Every conference, platform, and consultancy seems to have a view on how artificial intelligence will “transform delivery.” Dashboards will think for us. Schedules will self-correct. Risks will be predicted before they appear. Project managers, we’re told, will finally be freed from the grind.

There’s some truth in that. There’s also a lot of exaggeration.

After working across complex infrastructure, tourism, events, and high-risk delivery environments, I’ve learned to be cautious of anything promising transformation without accountability. AI is a powerful tool. It is not a substitute for judgment, leadership, or decision-making under pressure. And in project environments, those are the things that matter most.

What AI is actually good at in project management

AI excels at pattern recognition. It can process volumes of data that humans simply can’t. In the right environments, it can surface trends, highlight anomalies, and automate low-value tasks.

Used well, AI can:

  • Analyse historic schedules to identify recurring delay patterns

  • Flag emerging risks based on early signals in cost, time, or resource data

  • Automate reporting, freeing teams from manual updates

  • Support forecasting by modelling scenarios faster than traditional tools

These are real benefits. In data-rich, repeatable environments, they can meaningfully improve efficiency.

But efficiency is not delivery.

Where the hype runs ahead of reality

Most project failures don’t occur because teams lacked data. They occur because decisions weren’t made, trade-offs weren’t owned, or uncomfortable truths were avoided.

AI does not fix that.

AI can tell you a schedule is slipping. It cannot decide which scope to cut.
AI can highlight a risk trend. It cannot take accountability for accepting it.
AI can model outcomes. It cannot choose between political, commercial, and operational consequences.

The hype suggests AI will remove friction from project delivery. In reality, friction is where leadership is required. Removing it entirely would remove the role of the project manager itself — and that’s not how complex projects actually work.

The danger of confusing insight with action

One of the biggest risks with AI-enabled tools is the illusion of control. When dashboards become smarter and analytics more sophisticated, it’s easy to believe that insight equals action.

It doesn’t.

I’ve seen projects with exceptional reporting fail spectacularly because no one acted on what the data was clearly saying. AI doesn’t change that dynamic. In some cases, it can make it worse by overwhelming teams with signals that no one is empowered to respond to.

Information without authority is noise.

AI won’t fix weak governance or blurred accountability

AI is often positioned as a way to “support decision-making.” That’s true — but only if decision-making authority already exists.

If governance is weak, AI will simply expose it faster.
If accountability is unclear, AI will generate insights no one owns.
If leadership avoids hard calls, AI will become another reporting layer rather than a catalyst for action.

Technology amplifies systems. It doesn’t replace them.

Where AI genuinely adds value — today

The most effective use of AI in project management right now is not replacement, but augmentation.

AI works best when it:

  • Supports experienced project leaders rather than replacing them

  • Automates low-value, high-volume tasks

  • Surfaces insights that prompt better questions, not automatic decisions

  • Is embedded into existing governance and accountability frameworks

In these contexts, AI becomes a force multiplier. It sharpens judgment rather than attempting to replace it.

The skills AI can’t replicate

For all the excitement around automation, there are core project management skills AI cannot replicate — and likely never will.

Reading a room.
Navigating politics.
Making decisions with incomplete information.
Earning trust under pressure.
Standing behind a call when it’s unpopular.

These are not algorithmic problems. They are human ones.

The irony is that as AI becomes more prevalent, these skills become more valuable, not less.

The future of AI in project delivery

AI will continue to evolve. Tools will improve. Capabilities will expand. Project management will change — but not in the way the hype suggests.

The future is not AI-led projects. It’s AI-assisted leadership.

Projects will still succeed or fail based on clarity, accountability, and decision-making. AI can support those things. It cannot replace them.

The real risk is not that AI will take over project management. It’s that organisations will invest in technology while avoiding the harder work of fixing governance, accountability, and leadership.

And no algorithm can solve that.

Comments

Post a Comment

Popular posts from this blog

Ben Webb on Project Management: The Moment Projects Are Actually Won or Lost

  Projects are rarely won at handover. They’re won much earlier — often before construction starts, before systems are built, before the first dollar is spent. They’re won (or lost) in the first set of decisions . Early decisions carry disproportionate weight The earliest choices lock in: Risk exposure Delivery flexibility Commercial leverage Stakeholder alignment Once those decisions are made — or avoided — everything downstream becomes harder to change. Late-stage heroics get attention. Early-stage clarity saves projects. Why early decisions get delayed At the start, everything feels possible. There’s optimism. Optionality. A desire not to “close doors”. So teams delay commitment. They keep options open. They push decisions downstream. What they’re really doing is exporting risk to the future , where it’s more expensive, more visible, and harder to manage. What experienced delivery leaders do They accept a simple truth: You never have perfect information. So they: Decide with wha...
Read more

Ben Webb on Project Management: Why Most Projects Don’t Fail — They Drift

Image
  Projects rarely explode. They don’t collapse in a single moment of chaos. They  drift . I’m Ben Webb, an Australian project manager who has delivered complex infrastructure, tourism and public-facing projects under intense scrutiny — including Sydney Zoo, ICC Sydney and Sydney Harbour BridgeClimb. What I’ve seen repeatedly is this: Projects don’t fail because people don’t care. They fail because  no one forces decisions early enough . Drift is the most expensive risk in projects Drift shows up quietly: Decisions deferred “until the next meeting” Risks acknowledged but not owned Interfaces assumed instead of resolved Governance present on paper, absent in reality Schedules still exist. Reports still get written. But control slowly disappears. By the time leadership reacts, the cost isn’t time — it’s credibility. Planning doesn’t stop drift. Decisions do. Most teams respond to uncertainty by adding process : More reporting More reviews More layers of approval None of that...
Read more

Ben Webb on Project Management: Busy Is Not the Same as Progress

  Every project has that moment. The calendar is full. The inbox is overflowing. Meetings run back-to-back. And yet — nothing meaningful is moving. I’ve seen this on projects worth millions, sometimes billions. Teams working flat out, exhausted, committed… and completely stuck. Because activity is easy . Progress is not. Why projects love to stay busy Busyness creates comfort. It looks like effort. It feels like momentum. But it often hides the truth: Decisions aren’t being made Risks aren’t being confronted Accountability is blurred Direction is assumed, not agreed When progress stalls, teams instinctively add more : More workshops. More reporting. More coordination meetings. Very rarely do they stop and ask the uncomfortable question: What decision are we avoiding right now? Progress always looks quieter than you expect Real progress doesn’t announce itself. It looks like: One hard decision made early One scope line drawn clearly One risk owned, not “noted” One escalation done pr...
Read more