How the travel industry is actually using AI in 2026

Nathan Whittingham, Consultant at KAE, offers an overview of how the travel industry is using AI in 2026.

There is a particular kind of industry transformation that happens not with a bang, but with accumulation. AI in travel is that kind of change. Slowly, then all at once, it has moved from an experimental feature to operationally essential, reshaping how bookings are made, how planes are maintained, how hotels price their rooms, and how travellers navigate the world.

At KAE, we help shape the commercial strategies of banks, payment companies, and fintechs. Travel sits at an intersection that matters enormously to our clients, namely high transaction volumes, complex personalisation demands, fraud risk, loyalty economics, and a customer base that expects everything instantly. That crossover gives us an interesting perspective on the travel industry and how it’s changing.

What follows is a view across the travel value chain, grounded in examples that reflect where the industry actually stands in 2026.

Travel intermediaries: conversion, retention and the end of generic search

AI has shifted the commercial ground for Online Travel Agencies (OTAs) in a fundamental way. Google's AI Overviews now generate synthesised travel answers directly in search results, intercepting the queries that once drove OTA traffic. The same capabilities that intermediaries are deploying to improve conversion are being turned on them by their primary traffic source.

The response, for the most ambitious players, is agentic AI: systems that do not merely recommend but act, searching, comparing, booking, and confirming autonomously on the traveller's behalf. The move from 'here are your options' to 'I have booked it' is the most significant structural shift in travel commerce since online booking itself.

Trip.com's TripGenie: three years of evidence

In March 2026, Trip.com published three years of real-world TripGenie usage data, and the numbers are instructive. AI-assisted order volume grew approximately 400% year-on-year. Use of core tools including hotel comparison, menu assistance, and live translation grew around 300%. Nearly 60% of all interactions are now booking-related; not research or browsing.

What is perhaps more interesting than the volume figures is the behavioural data underneath them. Travellers from Hong Kong and Singapore use TripGenie as a real-time decision companion, booking hotels and navigating in-destination on the fly. UK and German travellers engage weeks in advance, working through itineraries methodically to reduce uncertainty. Among users of the hotel comparison feature, more than half chose the hotel the AI recommended.

That level of trust takes years of consistent, high-quality outputs to build. It is also a useful principle for any organisation building customer-facing AI: the goal is not to create one experience, but a platform capable of meeting diverse customers where they are.

Miso: agentic booking as native architecture

Miso represents the logical next step in the agentic travel stack: a purpose-built AI booking agent designed from first principles rather than retrofitted onto a legacy OTA model. Where conversational AI presents options, Miso completes the transaction, handling the full trip assembly cycle autonomously, from intent capture through to booking confirmation, via a simple text message.

The commercial model illuminates something important about where agentic travel is heading. The value proposition is not a better search interface; it is the elimination of the search interface altogether. A new distribution layer is forming that sits above both OTAs and suppliers, and whoever controls that layer controls the customer relationship.

Travel suppliers: operational efficiency at the engine room

For airlines and hotels, AI's most material impact has so far been in operations, specifically in areas where the cost of getting things wrong is high and the data available to predict failure is rich. What distinguishes the leaders is not just the technology they deploy, but the organisational commitment behind it.

Lufthansa Technik: maintenance intelligence as a commercial product

Lufthansa Technik's predictive maintenance AI, delivered through its AVIATAR platform, is a case study in how operational AI capability becomes a commercial product. The system processes operational data to forecast component failure, optimise shop visit scheduling, and convert unplanned maintenance events into scheduled ones. Lufthansa Technik's workforce of 22,000 now serves more than 800 airline customers and 4,500 aircraft under exclusive contracts worldwide.

The commercial significance extends beyond internal efficiency. Airlines that cannot justify building their own AI maintenance capability can purchase the outcome instead. In December 2025, Frontier Airlines added AVIATAR's Predictive Health Analytics, Condition Monitoring, and AI-based Technical Repetitives Examination modules to its existing deployment, a pattern of progressive adoption that mirrors how the best enterprise AI platforms scale.

The broader lesson is that AI-derived insight is itself a product. Lufthansa Technik's monetisation of predictive intelligence is a model increasingly being replicated across the travel industry.

Accor Hotels: AI revenue management across 5,800 properties

Accor's partnership with IDeaS to deploy the G3 Revenue Management System across its global portfolio is a case study in enterprise-scale AI adoption. The platform is now embedded as part of Accor's brand standards, with over half of its 5,836 hotels actively using it to price rooms automatically, integrating booking patterns, demand signals, competitor pricing, and market conditions in real time.

In 2025, Accor reported a 4.2% increase in RevPAR across its portfolio, with recurring EBITDA reaching a record EUR 1,201 million, up 13.3% at constant currency. The Obvio Hotels franchisee group offers a more granular illustration: deploying IDeaS across its Accor-branded properties in France and Switzerland, Obvio achieved a 14% increase in Average Daily Rate and a 9.5% improvement in Revenue Generation Index, by replacing manual spreadsheet-based forecasting with AI-driven analytics.

The pattern here is consistent with what we see in financial services: AI’s commercial returns compound over time, and the organisations that invested in clean data infrastructure first are now extracting disproportionate value from it.

Technology providers: rebuilding the infrastructure layer

GDS providers occupy an unusual position in the travel ecosystem. They are the infrastructure on which the industry runs, and yet they have historically been most constrained by the legacy systems they built. AI is the forcing function breaking this open.

Amadeus Nevio: modern airline retailing goes live

Amadeus Nevio is one of the most significant restructurings of airline retailing infrastructure in a generation. The cloud-native, modular platform helps carriers move from legacy Passenger Service System architecture to a dynamic, offer-and-order-based model aligned with IATA's Modern Airline Retailing standards, enabling personalised, real-time offers tailored to each traveller's context, preferences and loyalty status.

Adoption across 2025 and into 2026 has been substantial. Finnair became the first airline globally to create a Native Order powered by Nevio. Air France-KLM, Lufthansa Group across all nine of its airlines, and British Airways on Dynamic Offer Pricing have all followed. Amadeus reported FY 2025 group revenue of USD 6.5 billion, up 8.5% at constant currency, with its AI-native Nevio portfolio explicitly cited as a growth driver.

The platform's modularity is as important as its capability. Airlines can adopt components progressively rather than face a high-risk migration. In an industry where legacy dependency has historically been the biggest barrier to innovation, that is not a small thing.

Four challenges that cannot be solved by technology alone

Acknowledging the opportunity honestly means acknowledging the friction. Based on our research across sectors, four challenges emerge consistently, and in each case the solution is organisational as much as technical.

• Accuracy and trust: AI failures in high-stakes contexts (e.g., bad pricing, misleading recommendations, failed biometric identification) erode confidence quickly and are hard to recover from. The practical response is staged deployment: pilot in low-stakes contexts, build a track record, then extend into higher-stakes decisions. Trust is earned incrementally.
• Governance and regulation: the EU AI Act is soon to be in full enforcement. Travel companies processing biometric data or deploying AI in high-stakes customer decisions are within scope. Governance is most effective when it is designed in from the start rather than retrofitted. The organisations treating compliance as a commercial differentiator are ahead.
• Data infrastructure: AI is only as good as the data feeding it. Legacy system fragmentation is endemic in travel, and cleaning, structuring, and governing data at scale cannot be contracted out or accelerated past a certain pace. The investment case for data infrastructure is often hard to make in the short term; the cost of not making it becomes clear later.
• Organisational readiness: top-down commitment must be matched by bottom-up fluency. AI capability that sits in a single team does not scale. The gap between organisations that have distributed AI competence across functions and those that have not is widening, and it shows up in both deployment speed and outcomes quality.

What comes next

The near-term trajectory for AI in travel is fairly clear, and it runs through agency. The shift from AI as a recommendation engine to AI as an autonomous actor, making bookings, handling disruptions, optimising loyalty redemptions, managing itinerary changes, is already underway. The TripGenie and Miso examples point in the same direction: successful interfaces remove friction entirely, rather than reducing it.

Beyond agentic booking, the next wave is personalisation at the infrastructure level. Amadeus Nevio is the early expression of this: retailing built around the individual traveller's context rather than a standardised fare class. As that capability matures, the distinction between 'searching for a flight' and 'receiving an offer calibrated to you' will erode entirely.

For those working in adjacent sectors, payments and banking in particular, the travel industry is a useful leading indicator. The agentic AI experiments, the loyalty economics, the infrastructure transformation: these are arriving in financial services with the same force and the organisations paying attention now will be better positioned when they do.

This article is part of The Paypers’ Travel Series, which includes contributions on topics spanning emerging trends in travel payments, fraud and security challenges, regulatory and tax impacts, risk management and forex, as well as sustainability in the travel industry. For a complete overview of all the contributions featured, click here.

About the author

Nathan has extensive experience across banking, payments, and business technology, advising FTSE 100 and global institutions on pricing strategy and optimisation, proposition development, and go-to-market strategy. He has supported multi-market projects across the financial services industry, with expertise in competitive positioning, customer experience, and growth strategy. 

 

 

About KAE

KAE helps banks, payment companies and fintechs around the world make data-backed marketing, product and strategy decisions by uncovering deep customer, market and competitor insights. For over 30 years they have provided market-leading businesses with the customer and market intelligence needed to shape and execute go-to-market strategies, develop products and experiences, optimise pricing and communications, and build effective sales and partnership ecosystems.