OpenAI Operator, Anthropic Computer Use, Perplexity agentic search, and dedicated travel agents are routing travelers to seats — and the airlines that aren't in the answer aren't in the booking.
The next distribution channel for commercial aviation is the AI agent. OpenAI's Operator, Anthropic's Computer Use, Perplexity's agentic search, Google's agentic Gemini, plus a wave of dedicated AI travel agents — Mindtrip, Wonderplan, Layla, Vacay, Otto — are starting to book flights for real travelers. Some browse the airline's own site. Some pull from OTAs. Some use the airline's API. All of them make a recommendation before the booking happens — and that recommendation is built from the airline's citation footprint.
The airlines that show up in those recommendations win the booking. The airlines that don't lose share without ever knowing why. This is the playbook for showing up.
How AI Agents Actually Choose an Airline
Five inputs, in roughly this order:
1. The recommendation surface. When a user asks an AI agent "book me the best flight from New York to Tel Aviv next Tuesday", the agent first generates a recommendation. That recommendation is built from the same citation graph that powers the underlying AI engine's answer to "best airline from New York to Tel Aviv." Citation share on the recommendation prompt = inclusion in the booking flow.
2. Schedule and pricing visibility. The agent then surfaces actual flights, pricing, and seat availability. Carriers visible through GDS, NDC, and direct-API channels show up. Carriers missing from those channels don't.
3. Loyalty and account integration. Increasingly, AI agents respect the user's loyalty preferences. If the user has Mileage Plus status, the agent prioritizes United. The carriers that make loyalty data available to AI agents win incremental share.
4. Recent reviews and reputation signals. Some agents weight recent reviews, recent news cycles, and on-time performance data. Carriers with negative crisis citation footprints get downranked.
5. Direct deals and partnership integrations. Emerging: airlines that integrate directly with specific AI agents (API partnerships, NDC integrations, agent-friendly content) get preferential placement. This will become a meaningful channel.
The Five Strategic Moves for Airlines
1. Win the recommendation prompt. Citation share on "best airline from [origin] to [destination]", "most reliable airline", "airline with best on-time performance", "airline with best business class" is the leading indicator of AI-agent booking share. Trade, creator, and consumer business coverage all feed this.
2. Make schedule and pricing data structured. Schema.org Flight, OfferCatalog, Offer markup on the airline's booking pages. Clean NDC data. API access for agents that integrate directly.
3. Build agent-friendly content. A traveler asking "what's the on-time performance of [airline]?" — what does the agent retrieve? Carriers with published DOT-aligned performance data on their own newsroom, schema-marked and dated, surface cleanly. Carriers without this lose to third-party aggregators.
4. Integrate with leading AI agents directly. This is early but moving fast. United, Delta, and JetBlue have all signaled partnership exploration with major AI agent platforms. The carriers that move first set the integration template.
5. Track agent-driven booking share. Once measurable, AI-agent booking share becomes a board-level metric. Today the measurement is rough. Within 24 months it will be standard.
The Distribution Stakes
Airline distribution costs are one of the largest line items in airline economics — GDS fees alone account for billions in industry costs annually. The shift to NDC was supposed to reduce those costs and improve content. AI agents may accelerate that shift dramatically.
Two scenarios:
Scenario 1: Agents work through existing distribution. AI agents query existing GDS and OTA channels, surface results, recommend. Distribution economics don't change much. Citation share matters but the booking flow is preserved.
Scenario 2: Agents work through direct airline integrations. AI agents partner directly with airlines, querying the airline's own systems, bypassing the GDS layer. Distribution economics change dramatically. The airlines that integrate first capture share.
Both scenarios are running in parallel. The communications strategy has to support both.
What This Means for Aviation Comms
The communications team is now a distribution lever. Earned media, trade research, creator coverage, and AI engine citation hygiene aren't just brand-building. They're inputs into the booking decision that AI agents make for real travelers.
The implication is structural. Airline marketing and communications budgets that historically targeted brand awareness and consumer search are now also targeting agent recommendation. The metric is citation share — measured by category, by engine, weekly. The team is a hybrid of comms, GEO, and increasingly product and engineering.
This is what "the AI Communications Firm" actually means in airline operations.
