Ready to Act, Not Ready to Implement

Every revolutionary technology arrives in a world optimised for what it replaces. The telephone arrived in a world built around the telegraph*. Communications, corporate structures, culture and consumer behaviour were all designed for the old technology. That infrastructure became the barrier to adopting the new.

AI in passenger disruption is arriving the same way. And at Grounded 2026, airline leaders told us exactly what's holding it back. Spoiler: it isn't technology. In fact, it rarely is.

Airlines are ready to let AI act

We asked leaders from more than 35 airlines a simple question: if the technology worked tomorrow, how much decision-making would you hand to AI during disruption?

• 39% would let AI act within predefined rules

• 28% would let it act on its own and escalate exceptions to people

• 26% wanted AI kept to recommendations, with people making the final call

• 7% would consider full autonomy for most passengers

That's 74% prepared to let AI take action, not just advise. For an industry with a reputation for caution, that's a big appetite. Airlines aren't resisting AI.

But not ready to receive it

Then we asked what's stopping them.

• 41% said systems and data

• 27% said organisational readiness

• 18% said trust and governance

• 10% said the business case

• 4% said regulation

Regulation came last. The business case barely registered. 86% pointed to readiness within their own airline: systems, data, organisation and governance.

The technology may be ready before the airline is. That's the tension this series explores.

What "not ready" actually looks like

Airline leaders were clear about what needs to change. Four examples stood out:

• "Centralised source of truth: all teams have the same information and data access."

• "Network operations data must feed directly into communications… since today the aircraft side often gets information before customer-facing departments."

• "Give the recovery team one view of each passenger's journey."

• "Change must be driven top-down; individual advocates alone cannot shift the model."

The common thread? The same information, at the same time, with someone accountable for it.

An AI acting during disruption needs exactly that. It needs to see what operations see, when operations see it. It needs the passenger's whole journey, not just one flight. And it needs someone to own the rules it follows.

The architecture is already shifting. Airlines may not have decided which decisions to automate or how much autonomy to allow, but those questions are already shaping the technology they buy. Just look at the RFPs written over the last 18 months.

Three questions for your airline

1. Do your operations and customer teams see the same information at the same time?

2. Is there one view of each passenger's journey that the recovery team can act on?

3. Who owns the rules an AI would follow during disruption?

If any answer is unclear, that's where to start. Not with the model.

Next in the series

Next, we'll explore what airline leaders would automate first, and why the safest place to start isn't where passenger outcomes are decided.

*See Carlota Perez on technological revolutions; Paul David and Brian Arthur on path dependence and lock-in.

Data: survey and workshop votes from leaders of 35+ airlines at Grounded 2026. Workshop topics were voted urgent, realistic, or controversial. The results show direction, not a market-wide survey. Workshop figures are counts of votes, not numbers of airlines.

Part of a series based on Matt Walker's article "Will the revolution be an evolution?", which first appeared in PhocusWire. Read the full article: https://www.plan3.aero/insights/will-the-revolution-be-an-evolution