AI Gains Acceptance in Business Travel, but Humans Still Matter

By Sophie Lindqvist

Eighty-six percent of respondents in an Engine survey said they would be comfortable letting AI book a routine trip, while 76% said they would be comfortable letting AI decide whether a trip was worth taking. Together, those answers frame AI as a potential booking and decision layer, but they measure stated comfort rather than transactions.

Engine, which describes itself as “the AI-native platform for intelligent travel,” produced the research with Human After All, a London-based design and research agency. The study examines AI in business travel and its value for businesses, drawing responses from business travelers, the general population and industry specialists.

That mix offers a useful directional signal, not a market forecast. The announcement gives no sample size, geography, fieldwork dates or methodology, and it reports no bookings, transaction values, commission changes or supplier deals.

Engine’s platform description is a company claim, not proof of distribution scale. The survey needs testing against usage data, filings where available and market sizing (an estimate of total buyers and spending) before it can support a revenue forecast.

What does stronger AI adoption look like inside companies?

Larger organizations lead the reported adoption curve: 79% report having built AI into business travel, compared with 69% of smaller organizations. Across the respondent base, 77% say AI has increased visibility and control over travel, while 71% say policy compliance has improved.

The stated payoff is time, not customization. Forty-two percent said AI saves them time, compared with just 2% who mentioned personalization. The practical takeaway for sellers is to prioritize tools that remove friction from booking and policy checks, because buyers want systems to do less work, not merely appear more sophisticated.

Will travelers hand disruptions to AI?

Most participants were willing to use AI when plans change: 88% said it had made dealing with a travel disruption easier. Yet 58% would prefer a human to handle a mid-trip problem, showing that acceptance of automation stops short of full delegation.

That pairing suggests a hybrid service model: software could identify, rebook or explain changes, while a person retains authority when stakes rise. Suppliers and booking platforms combining rapid rebooking with live support may protect revenue; transaction-only interfaces risk looking interchangeable to buyers. Automation can win the routine workflow without eliminating the service layer.

Are bigger budgets creating more demand?

Sixty percent of respondents say travel costs have gone up. The report also says four in 10 organizations report traveling more than a year ago, while nearly half report bigger travel budgets.

For sellers, those findings point to possible transaction growth, but not a measurable revenue forecast. No figures show how much spending increased, how many bookings resulted or whether higher costs reflect traveler behavior rather than supplier pricing. The commercial signal is volume pressure, although the survey offers no basis for sizing the gain.

What changes for travel sellers?

If AI can judge whether travel is worthwhile and complete routine bookings, control of the booking decision can shift from the traveler to the software making that recommendation. In that model, distribution (the route by which travel products reach buyers) depends on being readable by recommendation software, rather than only ranking well on consumer booking sites.

For travel sellers, that makes structured pricing, availability and corporate policy data commercially important. A supplier that cannot provide those inputs could be ruled out before a traveler sees the product. Conversely, platforms that prove accurate recommendations aligned with corporate travel rules could capture more of the booking decision and strengthen their share of corporate spend.

Neither claim is a measured result in Engine’s research. The announcement provides no information about Engine’s customer base, supplier connections, commission model, how often proposed trips become bookings or accuracy, so it cannot establish that Engine or any seller will gain share.

It supports a practical sales argument for machine-readable offers (offers structured so software can compare them) and dependable service, not a forecast of bookings. The most credible signal is that corporate travel buyers want faster, more controlled transactions, so sellers should expect distribution advantage to follow demonstrable booking accuracy and reliable human support, not AI branding alone.