At T2RL Engage 2026 in London, T2RL’s Richard Clarke used his annual Market Update keynote to name a trend he calls “escaped functionality”: airlines and third parties developing or integrating offer, order, settlement, and delivery capabilities outside their PSS vendor’s stack, one component at a time. It is a different kind of disruption than the industry usually talks about. Not a challenger vendor taking market share, but functionality quietly leaving the incumbent’s boundary while the incumbent’s position remains largely intact.
Clarke opened with a sober read on where the industry actually stands. The transition to Offer and Order “haven’t seen much action,” he said, and the “dial isn’t moving much” on the share of passengers boarded via Order and Settle Delivery. That framing matters for what follows: escaped functionality is not evidence that NDC has finally taken hold. It is evidence that airlines are starting to work around the infrastructure question entirely.
The concentration behind that infrastructure question is stark. According to T2RL’s own recap of the session, posted to the company’s LinkedIn account, three major suppliers control around 75 percent of global inventory management, leaving the remaining vendors to compete for the rest. That figure comes from T2RL’s social recap of its own keynote rather than a formal published report, so it should be read as company commentary rather than a cited market study. One attendee’s account of the same keynote put the concentration higher still, suggesting the number may shift depending on which market segment or dataset was referenced on stage. But the broader picture is clear: a small number of players have held most of the market for years.
This is the distinction worth sitting with.
Escaped functionality is not PSS replacement. An airline pursuing it is not ripping out its core system or switching vendors. It is building, or buying from a third party, the specific pieces, an offer engine here, a settlement layer there, that let it move faster than its PSS contract otherwise allows.
The vendor keeps the inventory management business. The airline keeps the PSS.
What changes is which system actually decides what a customer sees and how a sale gets completed.
Dominance and relevance start to come apart.
According to accounts from the session, several forces are pushing that apart faster. Artificial intelligence is making it faster and cheaper to build these capabilities independently. AI-simplified interfaces are lowering the technical barriers to integration that used to require deep vendor-specific expertise. Regulatory pressure is challenging bundled PSS models directly. And vendor charges for synchronization and duplication, long a quiet cost of staying inside the incumbent stack, are making independent builds look increasingly attractive by comparison.
Clarke also identified the airlines most likely to move first: carriers above 30 million passengers annually. That creates the possibility of a bifurcated market in which the largest airlines increasingly build around their incumbent systems while smaller carriers remain more dependent on the traditional vendor stack.
Clarke was careful not to overstate the pace of change. Token and inference costs will limit how much AI-assisted development airlines can realistically pursue, and incumbent vendors are not standing still either. They may respond by bundling their offerings more tightly rather than loosening their grip.
The result is not necessarily imminent vendor displacement. It is a question of contractual exposure.
And that warning centers on data ownership.
Clarke emphasized that airlines need explicit contractual rights to experiment and innovate. Without them, he cautioned, the industry risks accumulating shadow inventory and duplicated infrastructure: inefficient and costly outcomes that emerge quietly rather than through any single dramatic failure.
He committed to tracking escaped functionality and PSS migration patterns closely in the coming years, treating the trend as an ongoing market signal rather than a one-off observation.
For airlines currently negotiating or renewing PSS contracts, the practical test Clarke posed is a narrow one.
Does the contract allow experimentation with AI-driven integrations without financial penalty?
And does it guarantee access to the underlying data needed to make those experiments meaningful?
An airline that cannot answer yes to both is not really testing whether this shift applies to its business. It is waiting to find out on someone else’s timeline rather than its own.
T2RL’s own written Outlook for 2026 makes a version of the same argument. It frames contract flexibility with vendors and partners as essential precisely because airlines cannot predict which pressure, geopolitical, economic or technological, will hit next.
Clarke’s keynote warning was therefore not simply an isolated conference soundbite. It fits within T2RL’s broader published position on what airlines need from their technology contracts in an increasingly unpredictable environment.
The concentration of the PSS market may remain intact even as the role of the PSS changes.
That is what makes escaped functionality interesting.
The vendors can continue to own the infrastructure while airlines gradually stop treating that infrastructure as the boundary of what they are allowed to build.
The next disruption in airline technology may therefore not be a new PSS vendor taking the market away from the incumbents.
It may be airlines quietly deciding that they no longer need their PSS vendor to provide everything.
