AI as opportunity: leadership, culture and the future of project delivery

At nPlan’s recent AI Day, we hosted Tommy Clarke, Head of Programme Controls, and Carolyn le Roux, PMO Director, from Sizewell C for a candid discussion about how AI is being applied on one of Europe’s most consequential infrastructure programmes.

In the first blog in this three-part series, we focused on how Sizewell C is using AI to strengthen decision integrity - accelerating the controls loop, stress-testing assumptions and introducing structured challenge into governance.

But that conversation also went somewhere more human.

Adopting AI on a live nuclear megaproject is not just a technical decision. It is a leadership decision. It affects roles, expectations, culture and how difficult information travels through an organisation.

In this second blog, we have compiled the most revealing moments from that discussion so you can absorb the key leadership themes without watching the full stream.

"It's thrive or you'll stagnate"

The starting point was direct: is AI a threat, or an opportunity?

Two things stood out to me here.

First, Tommy reframes the issue as professional evolution. The shift he describes is not about replacing expertise, but about moving from task-based activity to outcome-based contribution. The future of project controls, in his view, belongs to SMEs who can interpret, challenge and add value in workshops, not just manage processes.

Second, Carolyn grounds that optimism in governance discipline. Opportunity does not remove risk. It requires policy, principles and structured risk assessment. Responsible AI is not an afterthought. It is a prerequisite.

The combination is important: cultural reassurance without governance is fragile; governance without opportunity leads to stagnation.

Communicating bad news - can AI help?

Dev then moved the conversation into more uncomfortable territory.

On megaprojects of this scale, difficult information is statistically inevitable. The question is not whether it will arise, but how it will be handled.

Carolyn follows the example of Sizewell C CEO Nigel Cann in challenging the premise of “bad news.” There is no good news or bad news - there is just news. That distinction is not semantic. It reflects a deliberate effort by Sizewell C’s leadership to create an environment of transparency, collective problem solving and accountability.

On complex megaprojects, information distortion often creeps in long before crises emerge. Strategic misrepresentation, optimism bias and delayed escalation are well-documented failure patterns. A culture that treats information neutrally and encourages early surfacing of issues is a structural defence against those dynamics.

Tommy then sharpens the governance perspective. Boards can accept risk utilisation. What they struggle with is surprise - particularly surprise that could have been anticipated with better analysis or better information.

That is where tooling becomes operationally important.

By accelerating options analysis and providing probabilistic forecasts across the schedule, nPlan's AI reduces the likelihood of the “car in the ditch” moment. Instead of presenting unchallengeable, last-minute shocks, the organisation can surface emerging risk earlier, compare pathways quickly and give leadership genuine thinking time.

Speed matters here. Not speed for its own sake, but speed in generating credible, digestible and defensible options. That changes the tone of executive conversations from reactive defence to structured decision-making.

Taken together, this suggests something important: Sizewell C is not relying on culture alone, nor on tooling alone. It is combining a leadership stance that welcomes news with analytical capability that reduces avoidable surprise.

That combination is what makes resilience credible.

Successful AI implementation - what would it look like in practice?

Dev closed by asking what success would look like - a year into the engagement with nPlan. Here's what Tommy and Carolyn had to say:

What they describe is not disruption for its own sake. It is disciplined maturation.

Forward-looking reporting. Fewer manual spreadsheets. More time in workshops where expertise is applied to real decisions.

Carolyn also signals a shift in organisational position. Success means the board and programme leaders rely on project controls for clear and accurate information. The function becomes central.

Tommy’s examples reinforce the practical side. Faster schedule quality checks. Faster comparison of alternative pathways. Quantified risk exposure differences between options. Months of traditional analysis compressed into timely, structured insight.

And critically, the tool is placed in the hands of practitioners. It is not confined to a narrow use case defined at the top. That reflects confidence in the team and reinforces the earlier theme of role evolution rather than role reduction.

Across all three clips, the pattern is consistent.

AI is not positioned as an autonomous decision-maker. It is positioned as an enabler of clearer judgement, earlier escalation and more resilient governance.

Preparing for the Inevitable

Complex infrastructure projects do not fail because risk exists. They fail because risk is obscured, softened or surfaced too late.

By fostering a culture where there is no such thing as bad news, only news, and by equipping teams with tools that accelerate probabilistic analysis and option comparison, Sizewell C is deliberately reducing the space in which distortion can take hold.

That combination - cultural openness and analytical speed - does not eliminate difficulty. It makes it manageable.

Leadership under pressure is not about avoiding tough conversations. It is about having them early, with evidence.

‍