A couple of years ago, if you’d told me I’d be spending so much time thinking about AI, I’d have assumed you had the wrong person. I’m a planner and have been for over 20 years. I came up through the world of council planning, policy, plan development, statutory frameworks, community outcomes, and then found myself leading ‘walkable urbanism’ and 20-minute neighbourhoods.

What changed wasn’t my identity as a planner; it was the nature of the work, and the pace of technological advancement around it.

For most of my career, I’ve watched talented, committed people spend an extraordinary amount of time on tasks that are not “planning” at all: chasing missing information, reformatting reports, repeating compliance checks, the list goes on and on. Planning complexity has kept rising, but the tools and workflows haven’t kept pace, and the result was predictable: frustration, delay, and professional judgement being crowded out by administration and rework.

That’s the context in which AI became interesting, not as a fad, but as a better way of working. And it’s why I set up Spero-ai: to bring AI into planning and development work in a way that is trustworthy, auditable, and grounded in how regulated decisions are made.

Spero-ai is led by experienced planners, not led by the tech, so domain expertise and deep sector understanding sit at the heart of everything we do.

AI is disrupting planning, not replacing planners

When people talk about AI disruption, it’s often framed as replacement: AI taking jobs, automating decisions, removing professionals from the loop. That’s not what I believe will happen, and it’s not what good adoption should look like.

What’s being disrupted is the economics of time in planning. AI is changing how quickly information can be pulled together, how consistently documents can be checked, and how rapidly first drafts can be produced, freeing up capacity for higher-value human work: judgement, interpretation, stakeholder conversations, negotiation, and design thinking.

Organisations don’t want to be left behind. They can see competitors moving faster, they can see teams experimenting quietly, and they can see that those who learn to use AI safely will compound an advantage over time. But they also want trusted partnerships and sensible guardrails, because, if not designed properly, AI can be confident and wrong. It can miss nuance, overlook a buried condition, misunderstand a local policy context, or produce language that doesn’t sound right.

The planner’s job in the AI era: define the problem, control the risks

In my view, the planner’s advantage is not technical; it’s contextual.

You don’t need to build AI tools to get value from AI. You need to know the problem you’re solving, the constraints that matter, and what “good” looks like in a regulated setting. AI performs best when it’s given a clear task, strong inputs, and explicit guardrails and rules, which is exactly how we already think when we brief a graduate, or a consultant.

Just as importantly, planners need to stay on top of what AI can and can’t do, because safe adoption depends on knowing where the boundary sits between “useful assistance” and “unacceptable risk”.

Right now, AI is strong at factual drafting, summarising, extracting, comparing, reformatting, generating checklists, and highlighting likely omissions. Where it still needs professional leadership is factual reliability, interpreting ambiguity, and applying local policy nuance. In other words: the human remains responsible for meaning, judgement, and accountability, and the technology needs to be implemented in a way that makes that responsibility easier to discharge, not harder. 

This is central to how Spero-ai operates: AI should support the professional, not replace them, and it should make checking and traceability easier, not harder.

Many have been burnt by tech, and that history is shaping AI adoption

Another theme I hear constantly is that many organisations have had poor experiences with technology procurement. The traditional model has often been: a vendor goes away, builds a product, sells it to you, implements it, and then leaves you with it. You’re left with a tool that doesn’t match your needs, requires workarounds, and locks you into a contract or platform for years.

That history matters. It creates scepticism, and honestly, it should. In planning, we can’t afford systems that add friction or reduce confidence, because the downstream cost is delay, dispute, and reputational damage.

A newer delivery approach: agile, integrated, fast, and tailored

One of the most encouraging changes is that AI enables a different way to deliver technology.

Instead of “rip and replace”, modern tools can sit between systems: integrating with what you already use, drawing from your existing documents and data, and fitting into workflows rather than forcing a wholesale rebuild. You can start with a single workflow, prove value quickly, and iterate from there.

This agile approach changes the cost and speed equation. Meaningful improvements can be delivered quickly, sometimes in weeks or months, without multi-year implementation programmes. And because it’s iterative, the solution can be tested, changed, and tailored as you learn what works for your team, your local context, and your risk settings.

That’s also why starting small isn’t a lack of ambition; it’s a smart path to scale.

A growing opportunity for planners

This is where the recruitment lens becomes important. AI is creating demand for professionals who can bridge between domain work and digital capability.

In my view, the planners who will thrive are those who can define problems clearly, understand risk and evidence, stay current on AI capabilities and limitations, and design workflows where AI improves quality rather than eroding it. That combination, deep sector expertise plus practical AI literacy, is going to become increasingly valuable.

Final thought

If you’re leading an organisation, the goal isn’t to just “use AI”. This is a fundamental shift in how work gets done, and it needs broad organisational and strategic thinking, with a plan, not just scattered experimentation.

A sensible approach is to start by defining the outcome you’re aiming for, and then test your way towards it through light, low-risk adoption.

What is the goal:

• Staff spending materially more time on high-value professional judgement?
• Faster turnaround with fewer rework loops?
• More consistent quality and compliance?
• Or freeing capacity to expand into work you’ve always wanted to do properly, better engagement, stronger strategic planning, improved urban design input, proactive place strategies, deeper due diligence, or more rigorous post-approval follow-through?

AI can help unlock that, but only if it’s treated as a capability: governance, skills, workflow design, and trusted delivery that integrates with the systems you already rely on. Start small, prove value, refine fast, and scale with intent.

And if you’re a planner considering your own path: you don’t need to build the technology. But you do need to understand what it can and can’t do, and you need the confidence to direct it with clarity. That combination will be hard to beat.

James Mant MPIA is a town planner and the CEO of Spero-ai. After two decades in planning, he now focuses on helping organisations adopt AI in a practical, low-risk way that keeps professional judgement and accountability at the centre.