Here is a question this article will not answer in the first paragraph: when an AI agent books your next business trip, whose interest is it actually protecting, yours or whichever platform made the comparison easiest?
Hold that question. A new Accenture Research report, Win the AI Agent, Win the Traveler, gives the industry its most detailed picture yet of how close that question is to becoming a daily operational reality rather than a thought experiment, and the answer it points toward should worry any travel brand that has spent decades assuming loyalty and habit were enough to keep a customer in place.
Being essential stopped meaning being irreplaceable
Every travel company has always counted on something to anchor the customer relationship. An airline has seats, a hotel has rooms, many trips simply cannot happen without them, and loyalty status built up over years makes switching away feel like walking away from money already earned. An online travel agency has range and a familiar interface, both of which make returning easier than exploring alternatives.
An AI agent does not carry any of that sunk cost. When a traveler asks one to plan and book a trip, it assembles options, loyalty value, payment methods and destination spending, and systematically compares reliability, redemption value and payment perks until it lands on the best available fit for that specific trip, not the brand the traveler has always used.
Choices that used to run on habit are becoming algorithmic, and the 600 senior travel executives Accenture surveyed in June 2026 already sense the shift: asked who will own the traveler relationship in the near future, they ranked AI-native platforms ahead of individual airlines and hotels, 21% to 19%.
An intermediary category that barely existed a few years ago is now expected to sit closer to the traveler than the brands actually running the trip.
Travelers will delegate, but only up to a point
The traveler side of Accenture’s research, drawn from 5,003 respondents across 12 markets, shows genuine appetite for this shift, with limits that matter enormously for how a travel brand should respond. Nearly two in three travelers use four or more separate platforms to book a single trip, 94% call the process moderately or very complex, and 66% believe that complexity has already caused them to miss a better option they would otherwise have found. Almost all respondents, 96%, said they would switch away from a preferred travel brand for a meaningfully easier and faster experience, which turns a clunky booking flow from an annoyance into a structural competitive risk that loyalty can no longer insure against.
That openness has a ceiling, though, and the ceiling sits exactly where the money and the risk concentrate. Forty percent of travelers are willing to fully delegate comparing travel options to an agent. Only 10% are willing to fully delegate the actual payment. Discovery and loyalty optimization are where travelers welcome the agent doing the work; autonomous payment is a trust problem still to be earned, not a feature any company can simply ship and expect adoption of.
A traveler might authorize an agent to monitor fares and book automatically within a preset budget, preferred airlines and a 24-hour cancellation window, but almost none want to hand over the card entirely.
The fragmentation the traveler feels is a symptom of the fragmentation inside the building
The complexity travelers experience as an interface problem is, underneath, a structural one, and no clever chat window fixes it alone. It comes from decades of building payments, loyalty, identity and customer service as separate systems that were never designed to talk to each other. Executives are candid about it: customer data remains disconnected across their own organization according to 87% of them, payments according to 86%, identity and profiles according to 79%, and loyalty according to 72%. Those four systems are precisely what an agent needs connected if it is going to act reliably on a traveler’s behalf, and when they stay siloed, a nicer-looking app does nothing to change what the agent can actually execute.
Payments show what that internal fragmentation costs in the moment it matters most. Executives describe their own payment operations as fragmented across channels, currencies and provider relationships, and travelers feel it at the exact point of purchase: 83% have walked away from a booking because of a payment problem, a failed card, too few payment methods, or a fee that surfaced late in the flow. Agents make that failure mode worse rather than better, because a machine booking at scale repeats the same broken step many times over.
Accenture’s own simulation work puts the stakes starkly: 86% of agent-mediated transactions are at risk of being abandoned or switched to a competitor when something breaks during the purchase, with travel named among the categories most exposed to that risk.
OpenAI already lived a small version of this. It launched Instant Checkout inside ChatGPT with real fanfare, only to quietly step back from it seven months later. In its own words, the company found that “the initial version of Instant Checkout did not offer the level of flexibility” it wanted to provide, and redirected purchases back to merchants’ own checkout flows rather than closing the loop natively, as OpenAI explained in its own product update, a retreat that TechCrunch also covered in detail.
If a company with OpenAI’s engineering resources could not make in-chat checkout for ordinary retail goods reliable enough to keep running, the execution bar for travel, with its pricing volatility, cancellation rules, multi-currency payments and post-booking servicing, sits meaningfully higher.
Loyalty carries its own version of that same fragmentation cost, and the dollar figures involved explain why airlines in particular have been slow to open that layer up to outside comparison.
Tiers, shifting point values and partner restrictions have long made rewards genuinely hard for a traveler to compare unassisted, which is precisely the complexity that kept people locked into a single program when they had to do the math themselves. An agent removes that friction instantly, comparing redemption value and partner options the moment it is asked to.
The financial stakes behind that shift are not abstract: Delta’s remuneration from its American Express co-brand relationship reached $8.2 billion in 2025, up 11% year over year, according to the airline’s own year-end results, a single data point that shows how much of an airline’s real profitability now runs through a loyalty ecosystem an agent could, in principle, reprice for the traveler in seconds.
I should disclose why I am reading this report closely
Two days before this report reached my inbox, my own company launched something built for exactly the problem it describes. At Travel Connect, VOLL’s annual industry event in São Paulo, our CEO Luciano Brandão announced VOLL Intelligence, an agentic AI layer that lets travelers converse with agents that remember their preferences and lets travel managers pull budget, supplier and spend analysis into their own AI workflow, as PanRotas reported from the launch. I have an obvious stake in the argument that follows, so read the rest with that fully in view.
Latin America is the gap in both surveys
Here is the part of Accenture’s methodology worth sitting with for longer than a footnote. The 5,003-traveler survey covered the United States, Canada, the United Kingdom, Germany, France, Spain, Italy, India, China, Japan, the United Arab Emirates and Saudi Arabia, twelve markets, zero of them in Latin America. The 600-executive survey spanned 17 countries across Asia Pacific, Europe, North America, the Middle East and Latin America combined, which means the region contributed a slice of a pool the report never breaks out on its own. Independent coverage of the same findings, from HotelNewsResource to La Tribune de l’Hôtellerie in France, repeats the global figures without ever isolating a Latin American data point, because there isn’t one to isolate.
That silence is not a minor omission. A region running its travel programs across more currencies and more regulatory regimes per unit of geography than the markets Accenture actually measured is arguably more exposed to the exact fragmentation this report describes, not less, and nobody has published the number that would let a travel manager in São Paulo or Bogotá compare their own program’s agent-readiness against a real regional baseline. The gap matters especially now: GBTA’s own economic-impact research, presented in São Paulo the same week VOLL launched its agentic tool, found corporate travel generated more than US$ 3.4 billion in business-travel-related revenue for the city alone in the most recent year measured, with 7.3% growth year over year. A market moving that much money deserves its own agent-readiness benchmark, not an extrapolation borrowed from Frankfurt or Riyadh.
What Accenture tells each type of company to fix first, and why the order matters
Accenture’s prescription is sequenced rather than a single checklist, and the sequencing is the useful part. Airlines, it argues, should start by unifying traveler identity, profile, preferences, entitlements, loyalty and transaction history across every touchpoint, since loyalty modernization, disruption handling and future agent-led servicing all depend on knowing who the traveler actually is first. Air India’s 2026 app relaunch, which brought booking and payments in-house rather than through third-party vendors, is the report’s example of that identity layer being built deliberately rather than inherited piecemeal.
Hotels face a different first move: making their own value legible to a machine rather than a browsing human.
Rates, member pricing, upgrade rules and cancellation terms typically sit in systems designed for people, not for an agent that needs to compare them programmatically, and a hotel that keeps that data locked in human-readable pages risks simply never appearing in an AI-native recommendation set. Ascott, the Singapore-based hospitality group, is cited as an early mover here, restructuring its systems so its own inventory surfaces directly inside leading AI platforms rather than waiting for a human to find its website first.
Online travel agencies, Accenture argues, already have the shortest path of the three, since aggregation is their existing business model. Their real exposure is different: if an AI-native platform becomes the traveler’s actual entry point, an OTA risks being reduced to a transaction processor while the relationship, the data and the upsell opportunity all migrate to the layer sitting above it.
The report’s suggested starting point is payment orchestration, a reliable, multi-currency, multi-method payment layer that gives an agent real visibility into what it can use and how, since whichever party clears that bar first is the one that keeps the customer relationship when the agent decides where to route the booking.
Answering the question this piece opened with
So, back to where this started. Whose interest does an AI agent protect when it books a trip? Right now, mostly the traveler’s, within a narrow band of decisions the traveler is actually willing to give up, discovery and comparison far more than final payment. But that protection depends entirely on the agent having clean, connected, accurate data to work from, and today, per the travel industry’s own executives, that data mostly is not connected.
An agent cannot protect an interest it cannot see clearly, and a travel company whose systems remain fragmented is not protected by its brand or its loyalty program when the agent making the comparison simply routes around it, without complaint, without a failed transaction anyone notices, just a traveler who ends up somewhere else.
The travel companies that treat this as a Northern-hemisphere problem to watch from a distance are making the same mistake the two surveys already made: assuming a gap in the data means a gap in the exposure. Neither is true, and the region that has gone unmeasured twice in one report is the one with the least excuse left to wait.
About me
I am an entrepreneur with over 20 years of experience at the intersection of tourism and technology. I am co-founder and Chief Business Officer of VOLL, the largest mobile-first corporate travel and expense management platform in Latin America, where I lead the commercial strategy behind our AI adoption.
A first-cohort graduate of The AI-Powered Organization at Stanford Graduate School of Business and a Marketing specialist from Fundação Dom Cabral, I serve on the Tourism Council of FecomércioSP and on the Executive Council of the Latin American Association of Corporate Events and Travel Management (Alagev). I write and speak about innovation, digital transformation, entrepreneurial leadership, and the future of corporate travel.