Data privacy concerns and doubts about the accuracy of AI-generated output are slowing the adoption of artificial intelligence across the business travel sector, according to FCM, the corporate travel management division of the Flight Centre Travel Group.

The disclosure matters commercially. Travel management companies have positioned AI as the central lever for cutting servicing costs and defending margins as corporate buyers press for lower transaction fees. If corporate clients and travelers distrust the technology — either because sensitive itinerary and payment data feeds into AI systems, or because the systems return unreliable content — the efficiency case that TMCs have built their technology roadmaps around weakens.

FCM's assessment puts two specific friction points at the center of the slowdown.

Privacy liability. Business travel itineraries combine personal identifiers, payment details, travel patterns and client meeting information. Routing that data through AI tools raises questions about where it is processed, who retains it and how it is protected under regimes such as GDPR. Corporate travel managers, already accountable for data handling in supplier contracts, have a direct incentive to hold back deployment until those questions are answered contractually rather than rhetorically.

Accuracy risk. AI systems are known to generate confident but incorrect output — wrong flight times, invented hotel details, misstated policy rules. In managed travel, an error that books a traveler onto the wrong routing or outside policy creates real costs: rebooking fees, duty-of-care exposure, out-of-policy spend that escapes negotiated rates. Buyers weigh those downside scenarios against the productivity gains vendors promise.

The stakes are unevenly distributed. Large corporates with in-house legal and security teams can vet AI tools before rollout. Smaller accounts — the mid-market segment TMCs compete for most aggressively on price — often lack that capacity, which means AI-driven servicing reaches them last or arrives without the trust infrastructure that would make corporate buyers comfortable relying on it.

For sellers of travel, the practical consequence is sequencing. Tools that reduce internal cost — fare search, itinerary assembly, post-trip reporting — face fewer objections than tools that face the traveler or touch booking confirmation. Distribution effects follow the same logic: an AI layer that recommends properties or routes will shape share only if buyers believe its outputs are accurate and its data handling is clean.

FCM's caution also functions as a market signal. When a TMC of its scale, inside a group with substantial technology investment, frames adoption as stalled rather than accelerating, it tempers the expectation that AI will rapidly displace agent servicing or compress the cost base of managed travel programs. Measured results, not vendor projections, will determine how fast that shift actually runs.

The question FCM's warning leaves open is temporal: whether privacy and accuracy concerns prove a temporary brake — resolved by better guardrails, audit trails and contractual clarity — or a structural constraint that keeps a human servicing layer central to corporate travel economics for years to come.