Amadeus Defends GDS Relevance as AI Agents Turn to Aggregator Alternatives for Travel Booking Infrastructure

The travel technology landscape is undergoing a profound structural shift as generative AI agents transition from experimental chatbots to functional booking engines. This evolution has sparked an intense debate regarding the future of Global Distribution Systems (GDS) like Amadeus, which have historically served as the essential connective tissue between airlines and travel retailers. During the Q3 2026 earnings call held on Friday, November 6, the leadership team at Amadeus addressed a growing strategic challenge: why emerging AI platforms, such as Meta’s Muse, are increasingly bypassing traditional GDS integrations in favor of modern, API-first aggregators like Duffel.

The Changing Architecture of Travel Distribution

For decades, the GDS model has been predicated on a centralized, highly complex ecosystem that manages vast amounts of data, inventory, and payment clearing processes. However, the rise of AI agents—which are designed to operate with speed, agility, and low latency—has exposed the friction points inherent in legacy booking systems. When an AI agent assists a user in planning a trip, it requires real-time access to flight inventory, ancillary services, and the capability to execute complex modifications.

The partnership between Meta’s Muse and the London-based travel startup Duffel serves as a microcosm of this broader shift. By utilizing Duffel, Meta’s AI agent gains direct access to airline content through a streamlined, developer-friendly interface that avoids the technical overhead associated with traditional GDS legacy protocols. This architectural preference poses a fundamental question: as the front end of travel shopping moves toward conversational AI, will the back end move away from the traditional GDS towards leaner, cloud-native aggregators?

The CFO’s Perspective: Validation Through Disruption

Addressing investor concerns during the November earnings call, Amadeus CFO Carol Borg maintained a posture of strategic confidence. Borg argued that the emergence of players like Duffel does not signal the obsolescence of the GDS model but rather its validation.

"The Muse-Duffel partnership supports our core thesis that AI platforms will prioritize the use of existing aggregators rather than attempting to build deep, complex booking and settlement systems from scratch," Borg stated. From the perspective of Amadeus, the core value proposition of a GDS—managing the intricacies of global airline distribution, complex fare filing, and regulatory compliance—is too burdensome for AI startups to internalize.

Borg’s assessment suggests that while AI platforms may choose different entry points for their data, the underlying infrastructure requirements remain consistent with the services Amadeus provides. In this view, if a company like Duffel succeeds in facilitating a high volume of bookings, they will eventually reach a scale where they must either integrate with a GDS or replicate the massive, multi-decade investment in infrastructure that Amadeus has already perfected.

A Chronology of the Shift

To understand the current tension, it is necessary to examine the evolution of travel distribution over the past decade:

• 2015–2018: The Rise of NDC: The International Air Transport Association (IATA) introduced the New Distribution Capability (NDC), aimed at modernizing airline retail. This initiative allowed airlines to offer more personalized content, bypassing some traditional GDS constraints.
• 2019–2021: The Pandemic Reset: As travel demand plummeted, the industry focused on digitizing operations. The shift toward contactless services and automated support became a survival imperative.
• 2022–2024: The Generative AI Boom: With the release of advanced Large Language Models (LLMs), companies like Meta, Google, and OpenAI began integrating travel planning capabilities directly into their interfaces.
• 2025–2026: The Aggregator Pivot: Companies like Duffel, RateHawk, and various specialized API providers began securing high-profile partnerships with AI firms, positioning themselves as the "API layer" between the AI agent and the airline inventory.

Supporting Data and Market Dynamics

The travel distribution market remains highly concentrated, with Amadeus, Sabre, and Travelport holding the majority of the market share. However, the segment of "modern aggregators" is growing at a compound annual growth rate (CAGR) that significantly outpaces the traditional legacy distribution segment.

While the total volume of bookings processed through traditional GDS remains in the hundreds of millions annually, the new segment of AI-driven, intent-based bookings is showing a preference for agility. Modern aggregators typically offer integration times measured in days or weeks, whereas legacy GDS integrations can take months, involving complex certification processes and technical debt.

Furthermore, the cost structure for AI agents is highly sensitive to latency. If an AI agent takes ten seconds to fetch a flight result due to a sluggish legacy interface, the user experience is compromised. Duffel’s ability to provide sub-second responses via modern APIs has made it a preferred choice for companies prioritizing user experience (UX) above all else.

Implications for the Travel Tech Ecosystem

The friction between traditional GDS providers and new-age aggregators represents a broader battle for control over the "travel intent" funnel. If the primary point of contact for a traveler becomes a chatbot like Muse, the entity that controls the connection to the airline inventory holds significant leverage.

If Amadeus and its competitors are to remain dominant, they must evolve their technical offerings to be as accessible as the startups currently threatening their niche. This involves more than just offering an API; it requires a fundamental change in the developer experience. Analysts have noted that Amadeus has been investing heavily in its "Amadeus for Developers" portal, yet the speed at which it can modernize its core systems is hampered by the need to maintain reliability for its massive existing client base of global airlines and travel agencies.

Potential Reactions from Industry Stakeholders

Industry experts have offered a range of reactions to the Muse-Duffel development. Some suggest that this is a classic "innovator’s dilemma," where the incumbent is distracted by the needs of its largest, most traditional customers, leaving the door open for smaller, faster players to capture the emerging AI market. Others argue that AI platforms are still in their infancy; once these agents scale to handle millions of bookings, they will find that the specialized, regulatory-heavy services provided by Amadeus—such as complex multi-leg rebooking, global tax handling, and multi-currency settlement—are not so easily replicated by lightweight aggregators.

Airlines themselves are caught in the middle. They want the reach that AI platforms provide, but they also want the stability and global reach of the GDS. Most major airlines are currently pursuing a "hybrid distribution strategy," maintaining their GDS contracts while simultaneously exploring direct-connect APIs or specialized aggregators to feed their inventory into new sales channels.

The Path Forward: Integration or Competition?

The long-term outlook for Amadeus hinges on its ability to position its vast data repository as an essential component for the AI era. If AI agents rely on aggregators, Amadeus may seek to become the primary data provider for those aggregators. By providing the deep, backend infrastructure that powers the likes of Duffel, Amadeus could effectively remain the backbone of the industry, even if its brand is not the one the AI agent interfaces with directly.

However, the risk remains that if AI agents become the primary gatekeepers of travel, they may seek to disintermediate the GDS entirely by forming direct partnerships with airlines. This "direct-connect" model is already popular among low-cost carriers and is increasingly being adopted by full-service legacy airlines eager to avoid GDS booking fees.

In conclusion, the situation with Muse and Duffel is more than a minor technical footnote; it is a clear indicator that the architecture of travel is being rewritten. As CFO Carol Borg noted, the GDS model is not necessarily being displaced, but it is certainly being challenged to prove its value in a world where speed, developer access, and AI-native architecture are the new currencies of the travel industry. Amadeus, with its deep institutional knowledge and massive scale, remains the giant in the room, but the room is getting smaller, faster, and much more automated. Whether Amadeus can successfully pivot to become the "API of record" for the AI age will define its performance for the next decade.