In May 2026, FLYR announced that Riyadh Air had become what it called the first full-service carrier to launch based exclusively on an Offer and Order framework, built on FLYR’s own platform rather than on a traditional GDS foundation. The claim comes from FLYR’s own press release, so it should be read as the vendor’s framing of its marquee customer rather than independently verified fact. But even allowing for that, the underlying architecture is real and well documented elsewhere: Riyadh Air is a digitally native airline that did not inherit a legacy GDS-centered stack the way every carrier before it effectively had to.
That single fact reframes a question the travel distribution industry has quietly avoided asking with any seriousness for two decades. Amadeus, Sabre and Travelport between them are estimated by the travel technology research firm AltexSoft to control roughly 97 percent of global GDS air bookings, based on air-booking volume rather than audited company disclosures. IATA’s own tracking of the industry shows just how early this transition still is: as of October 2025, only seven airlines had signed full-scope contracts to move to Offers and Orders, with 48 further pilots and proof-of-concepts underway. Against that backdrop, a digitally native carrier walking straight past the incumbents looks less like an anomaly and more like a signal. The more interesting question is not whether a fourth company will rise to challenge Amadeus, Sabre and Travelport head to head. It is whether the GDS itself, as a single bundle of functions, is already being taken apart.
The last challenger got bought, not beaten
The honest answer starts with a case most people in distribution remember only vaguely. In August 2019, the US Department of Justice sued to block Sabre’s proposed $360 million acquisition of Farelogix, a smaller, NDC-focused technology company that two major US carriers had testified only Farelogix could serve at the level they needed. The government’s complaint was blunt. It alleged Sabre had “operated outdated technology and resisted innovation” while Farelogix had “stepped in to address the needs of airlines,” and that folding Farelogix into Sabre would eliminate exactly the kind of competition that had, in the DOJ’s words, “substantially benefited airlines and consumers.”
Sabre initially won that fight in the United States. Judge Leonard Stark ruled in April 2020 that the government had failed to prove the deal would harm competition, closing out an eight day bench trial with judgment for the defendants; that judgment was later vacated as moot once the transaction itself was abandoned. Two days after the initial ruling, the UK’s Competition and Markets Authority reached the opposite conclusion and blocked the same merger outright following a Phase 2 investigation. Sabre and Farelogix had already spent roughly 20 million dollars in legal fees defending the deal in the US alone, and Sabre ultimately recorded a further 46 million dollar charge tied to the UK ruling before walking away from the acquisition entirely.
Two governments looked at the same merger and reached opposite verdicts on whether a genuine independent NDC challenger disappearing into Sabre was a competitive problem worth stopping. That ambiguity is itself the story. Farelogix was not a conventional fourth GDS in the sense of a company trying to build a rival marketplace; it was a narrower NDC and airline retailing technology challenger, and that distinction matters. But it is precisely because Farelogix stayed narrow, useful and dangerous only at the edges of what a GDS does, that it is one of the clearest test cases the industry has. It did not disappear because the market had definitively rejected it. It was caught in an acquisition attempt, contested in court, and ultimately survived as an independent company after regulators in the US and UK reached opposing conclusions. Nothing about that episode suggests the three-way structure is fragile against a direct challenger. What it suggests is that a narrow, sharp-edged threat aimed at one function of the GDS bundle is exactly the kind of thing an incumbent can move fast enough to try to buy.
Retailing, connectivity, distribution, shopping: someone is already picking each one off
A traditional GDS has historically bundled several functions into one distribution system: a single connection point across many carriers, aggregated shopping and availability, distribution into agencies and travel management companies, and increasingly the ability to consume and distribute richer airline content as carriers move toward NDC and Offer and Order. Rather than asking which single company might rebuild that whole bundle from scratch, a sharper way to read 2026 is to look at each function separately and ask who is peeling it off.
The retailing layer is where FLYR is furthest along. The company has raised more than 500 million dollars to date, including a 225 million dollar Series D round with participation from Avianca and additional credit facilities led by Vista Credit Partners. Its Offer and Order Management System is now live end to end at Riyadh Air, and in June 2026 the company extended its reach to the airport itself through a partnership with Res2’s iPort platform, covering check-in, baggage and boarding directly on FLYR’s own order data. That is a materially more complete retailing stack than anything the challenger category has produced before. But FLYR is not itself attempting to recreate the GDS marketplace. Its strength is the airline’s own retailing and order architecture, not becoming the place where agencies shop across hundreds of carriers at once.
The connectivity layer is where Duffel is making its case, and it is worth being precise about what kind of case that is. Duffel is not rebuilding the traditional GDS model on modern infrastructure. It is structurally different: a single, modern REST API spanning more than 300 airlines, built API-first for developers rather than around the agency terminal and PNR logic that Amadeus, Sabre and Travelport still carry. It is backed by Index Ventures and Blossom Capital, markets itself directly as the cleanest migration path for developers coming off Amadeus’s shrinking self-service API, which closed to new signups in July 2026, and already routes bookings for named corporate travel customers including TravelPerk. It has stated ambitions to build a “Universal Travel API” spanning rail, hotels and ground transport, with a public target of reaching 50 percent non-flight booking volume and eventual IPO readiness. Its proposition is therefore different from that of a conventional GDS: Duffel abstracts connectivity, including GDS and NDC sources alike, behind a single API rather than recreating the full agency distribution environment that incumbents have spent decades building.
The agency distribution layer is already partly unbundled in the markets TDN covers most closely. NDC aggregators including Verteil, TPConnects and AirGateway function as a genuine parallel distribution channel for carriers like FlySafair, Airlink and Ethiopian across Africa and the Gulf. They are solving a real connectivity problem the legacy GDS have been slow to address in these markets. None currently shows the capital base or stated ambition to become a full marketplace competing head to head with Amadeus, Sabre or Travelport. Their role today is translation layer, not challenger, but it is a layer the Big Three no longer fully control.
The shopping layer itself may be the one that matters most, and it is not a company at all. Several 2026 analyst notes, including IDC’s FutureScape hospitality and travel predictions and OAG’s Airline Tech Innovation Radar, argue that the real disintermediation pressure is not a new entrant but a new kind of buyer: the AI agent, sitting above all three incumbents and asking a blunt question none of them were built to answer, which is simply whether they can respond to an agent in real time rather than to a human at a terminal. One industry benchmark published in June 2026 found that only 11 percent of hospitality organizations had deployed an agent capable of completing a real-time booking at all. Sabre’s own Mindtrip and PayPal launch in May 2026 is best read as a defensive answer to that pressure, an attempt to make sure that when agentic booking does mature, Sabre’s rails sit underneath it. This is not a fourth candidate competing for the same prize as FLYR or Duffel. It is a force that could make the question of who owns each layer of the old GDS bundle irrelevant, by changing who or what is doing the shopping in the first place.
Nobody has to become Amadeus to make Amadeus less necessary
There probably will not be a fourth GDS in the traditional sense, a single company that rebuilds the whole bundle and sits alongside Amadeus, Sabre and Travelport as a fourth marketplace. That is the wrong shape of threat to watch for. The more serious pressure is that the market no longer needs any one company to perform all the functions a GDS historically bundled together. FLYR is peeling off retailing. Duffel and the NDC aggregators are peeling off connectivity and agency distribution. Agentic AI is quietly changing who does the shopping in the first place. None of these forces is trying to become Amadeus. Each is making it less necessary for anyone to be Amadeus, Sabre or Travelport in the way those three have defined the category for decades.
None of the three legacy GDS is being replaced wholesale in 2026. But none of them is competing on a single, stable definition of what a GDS actually is anymore either. That is the more interesting story, and probably the more durable one for readers to sit with. The Farelogix precedent suggests the incumbents have both the capital and the appetite to acquire their way out of any single narrow threat that gets too close. What none of them can acquire their way out of is a market that stops agreeing on what business they are actually in.
