The Importance of Independence in the Era of Offers and Orders Part 3
The Importance of Independence in the Era of Offers and Orders: Part 3 – AI's Role in Reshaping Airline Distribution
Guest post by Cory Garner , CEO of Garner
Revisiting the Case for Independence
In OpenJaw’s two previous articles (Part 1: https://www.openjawtech.com/insights/the-importance-of-and-path-to-airline-distribution-and-technology-freedom-1, Part 2: https://www.openjawtech.com/insights/part-2-the-importance-of-and-path-to-airline-distribution-and-technology-freedom-2), the case was made that legacy intermediaries like Global Distribution Systems (GDSs) tether airlines to outdated economics and limit their agility. Those pieces highlighted the need for carriers to break free to control their futures. Now, in mid-2025, as large language models (LLMs) and AI redefine how travellers discover and book, the case for independence is more critical than ever. In this guest piece, OpenJaw has asked me to explore how LLMs could emerge as the new chief intermediaries, connecting airlines directly to consumers while side lining GDSs and reshaping travel agencies’ roles. This shift amplifies the urgency for airlines to partner with GDS-independent technology providers for AI agent integration (via the emerging Multi-agent Collaboration Protocol, or MCP), NDC, and offer-order systems. Independence is no longer optional—it’s the key to thriving in an AI-driven world.
The Outdated Architecture of Airline Distribution
Airline distribution remains a double-intermediated relic. GDSs—Sabre, Amadeus, Travelport—aggregate supplier inventory and pipe it to agencies, handling roughly 40-50% of global bookings. Agencies, from online travel agencies (OTAs) like Booking.com to travel management companies (TMCs) like SAP Concur, layer on commissions or markups. Originally built on 1960s tech and still limping along with EDIFACT protocols, this model extracts 3-5% in GDS booking fees and commissions and fragments data, stifling personalization. NDC, with its XML-based standards for dynamic offers like ancillaries, aimed to modernize this, but adoption hovers below 30% for most carriers who have tried, hampered in part by GDS inertia.
AI and LLMs: A New Path Forward
AI, particularly LLMs like Grok or ChatGPT, is poised to upend this structure. Imagine a traveller saying, “Find me a flight from Miami to London with extra legroom and a vegan meal.” An LLM, leveraging the Multi-agent Collaboration Protocol (MCP)—an emerging standard for seamless AI agent interactions—could query airline APIs directly, assemble offers, and complete orders without GDS involvement. Airlines could partner with LLM providers, feeding inventory via MCP integrations to enable conversational bookings. Turkish Airlines’ 2025 MCP pilot (https://www.remote-mcp.com/servers/turkish-airlines) exemplifies this potential, testing AI-driven booking flows that integrate with their systems to deliver personalized offers, bypassing traditional GDS channels. This initiative, building on their 2024 GDS surcharge strategy, signals a shift toward direct, AI-enabled distribution. If LLMs become the primary point of content aggregation and booking management, GDSs lose their raison d’être. Why rely on legacy aggregation when AI agents, coordinated via MCP, can dynamically pull from sources?
The Declining Relevance of GDSs
This shift threatens to erode GDSs’ economic foundations. Their revenue stream depends on being the indispensable middleman, but LLMs excel at synthesizing data, navigating APIs, and delivering context-aware results. As noted in OpenJaw’s previous pieces, GDSs are chained to rigid tech that struggles with AI’s fluidity. GDSs won’t disappear overnight—their entrenchment in legacy TMC architectures keeps them afloat—but their dominance fades as AI redistributes value to airlines and LLM platforms. Airlines must act swiftly to avoid being tethered to a declining model as AI accelerates the shift to direct connections.
Travel Agencies: Adapting to a New Role
Travel agencies face a divergent future. Some will adapt, but not as traditional “agents” in the supplier-customer transaction flow. Instead, they’ll pivot to customer-side service providers with responsibilities delegated from the customer only. In the current model, agencies profit from commissions or GDS-fed bookings. In an AI world, LLMs handle transactions directly via MCP integrations, but agencies could gain delegated access to booking data to manage itineraries, handle disruptions, or curate experiences. For leisure travellers, niche agencies—luxury planners or adventure specialists—might enhance LLM outputs with tailored add-ons like local guides or exclusive experiences. In corporate travel, TMCs could oversee compliance or expense integration with enterprise AI systems, ensuring policy adherence while streamlining processes. Agencies of all kinds will gradually become less dependent on supplier-side compensation and more dependent upon customer-side compensation. The catch? Agencies tied to GDS economics will face severe pressure as LLMs take over search and booking, leaving them scrambling to redefine their value in a rapidly changing landscape.
The Imperative of GDS-Independent Technology
This brings me to the core argument: airlines must prioritize GDS-independent technology providers for MCP integrations, NDC implementations, and offer-order systems. MCP, as the emerging standard for AI agent collaboration, enables seamless communication between airline systems and LLMs, ensuring dynamic, personalized offers across channels. Offer-order tech, underpinning NDC and real-time pricing, demands flexibility to integrate with these AI ecosystems. GDS-affiliated vendors, however, often prioritize their parent’s ecosystem, embedding biases that favour legacy flows over direct, AI-driven connections. This misalignment stifles innovation and locks airlines into unfavourable economics, undermining their ability to leverage AI’s full potential. By choosing independent providers, airlines can ensure their technology stack supports the fluidity and personalization that LLMs and MCP enable, positioning them to capitalize on the shift to direct distribution.
Capturing AI’s Potential Through Independence
Why emphasize independence so heavily? It’s about capturing AI’s upside. LLMs, via MCP, enable hyper-personalized offers—think real-time ancillaries like priority boarding or carbon offsets—and seamless customer experiences, like predictive rebooking during disruptions. But these benefits demand tech unencumbered by GDS ties. Affiliated systems may limit MCP integrations or prioritize legacy fees, ceding control to old intermediaries just as new ones emerge. Independents give airlines the agility to navigate this shift, echoing my earlier calls to strategically cycle intermediaries. This approach ensures carriers can adapt to AI’s rapid evolution without being weighed down by legacy conflicts, allowing them to retain control over data, pricing, and customer relationships in an increasingly competitive market.
Navigating the Challenges Ahead
Challenges loom. Delegated access for agencies raises privacy concerns, requiring robust consent frameworks to protect traveller data. LLMs can falter—hallucinations or biases could undermine trust, necessitating human oversight to maintain reliability. Some airlines may opt to offer AI-enabled experiences in their own websites and apps only, depriving the larger AI ecosystem of better content and functionality, and themselves of improved revenue and customer satisfaction. Yet, the momentum is undeniable: AI is tilting distribution toward direct, customer-centric models, and independence is the linchpin for airlines to stay competitive. Carriers must weigh these factors carefully, balancing innovation with operational stability to fully harness AI’s transformative potential.
A New Era of Distribution
In closing, as LLMs rise as powerful intermediaries, we’re not ridding airlines of “800-pound gorillas” that dominate distribution—just cycling them from stage left to stage right and producing an ROI in the process. The AI-enabled future, powered by MCP and direct connects, offers airlines unmatched revenue potential through dynamic offers and unparalleled customer convenience. These new gorillas are preferable to the old, but only if airlines choose GDS-independent providers for MCP, NDC, and offer-order tech. As I’ve stressed across this piece, independence isn’t just a strategy—it’s existential.