Why we need to stop talking about AI pilots

Every airport we speak with has an AI pilot running somewhere. Predictive maintenance. Passenger flow forecasting. Security image analysis. Fewer can say, with real evidence, that any of it has changed an operational outcome. 

That’s the conversation the industry needs to have more honestly. There has been plenty of ambition around AI and digital transformation, and comparatively little scrutiny of results. We talk about digital twins and agent-based AI at conferences, then return to airports where the underlying data still lives in three systems that don’t talk to each other. 

The technology has genuinely moved. AI is no longer just producing recommendations for someone to act on; some systems now act directly, inside closed loops, at speed.  

That’s a meaningful shift, and it’s exactly why governance has become the harder problem. When a human made the call, accountability was obvious. When an algorithm does, most organisations are discovering they were never built to answer who owns that decision, what evidence backs it, or how to explain it to a regulator, an airline, or a passenger affected by it. 

The answer isn’t to slow down. It’s to be far more disciplined about proof before scaling, and far more deliberate about governance before it’s needed, not after an incident forces the question. 

This is exactly why the roundtable format matters more than another stage presentation. Nobody in the room is selling anything. It’s airport leaders comparing what’s actually worked, what quietly failed, and what they’d do differently, with peers who carry the same accountability. 

If you’re setting AI or data strategy for your airport in the next year, this is where that thinking gets tested. Bring the deployment that didn’t work as well as the one that did. That’s where the real learning happens. 

Register your interest in the IAS 2026 Roundtables on Digital Transformation, Data & AI. 

Register today