Airlines Use AI to Price Seats and Keep Travelers Close
The price of a seat, the value of a mile and the explanation for a delay are becoming parts of the same airline AI strategy. Carriers are using artificial intelligence to make daily decisions more precise and more personal, moving it from the fare screen to the loyalty account and, when a flight goes wrong, the passenger’s phone
Pricing is the most sensitive test. Airlines have adjusted fares according to demand for decades, but AI can analyze more variables and recommend changes far faster. In an August 2025 letter addressing lawmakers’ concerns, Delta said its AI pricing pilot uses aggregated purchasing data, route demand, competitive offers and operating costs. The airline said it doesn’t use personal information to set individualized fares, doesn’t share personal data with technology partner Fetcherr and keeps pricing analysts in control of the recommendations.
Other carriers are moving in the same direction. Thai Airways said in February that it would adopt Amadeus tools for AI-driven pricing and personalized offers across sales channels. Amadeus said its pricing product adjusts fares in real time using booking context and seat availability and can produce an average revenue increase of more than 3%
The transition may be less like turning over the controls than adding a faster co-pilot. Amadeus’ chief scientist wrote in March that hybrid forecasting models, which combine established airline pricing methods with artificial intelligence, consistently beat both traditional models and AI-only versions in tests. Pure AI models can miss basic business relationships, he said, including the normal link between higher prices and lower demand
Loyalty offers the next opportunity and a complication. Delta Concierge, an AI assistant available in beta to SkyMiles members, gives real-time answers inside the airline’s app and draws on trip details and customer preferences. Delta said in May that it planned to make the assistant available to all eligible customers by the end of summer. By reserving the tool for SkyMiles members, Delta turns convenience into a reason to enroll and gives the airline another direct channel for personalized offers.
AI outside an airline’s app could weaken that advantage. Accenture reported in June that 37% of loyal consumers would let an artificial intelligence agent switch from a favored brand when it found a better fit. Airline programs built around habit and accumulated miles could face a new intermediary that compares the fare, schedule and benefits without sentiment
Customer communications provide a more immediate payoff. United says generative AI gathers operational data and drafts plain-language delay messages for its “Every Flight Has a Story” program, with human employees reviewing the information. In June, CEO Scott Kirby described a goal of eventually explaining every delay through automated texts and even maintenance videos without human intervention. Better explanations can reduce uncertainty during disruptions and, Kirby argued, create more loyal customers.
Together, these projects show AI becoming a commercial layer across the trip. The same technology can help price the seat, personalize the relationship and explain why the plane is late. That creates room for airlines to sell more intelligently and serve passengers better. It also raises the cost of a bad recommendation. The carriers that gain the most will make AI useful where travelers can see it, explainable where money is involved and correctable when the system gets something wrong
