Australian travelers resist AI hotel booking, SiteMinder finds

By Elena Vasquez

Only 3.6% of Australians prefer an AI assistant or agent as the method for booking their next hotel stay, while 44% would rather book with no AI help, according to SiteMinder’s Changing Traveller Report 2027.

Australia ranks fourth-highest among 15 markets surveyed for preferring no AI assistance, behind the UK at 56%, Germany and Canada at 46% each. Respondents in India, Indonesia and Thailand register only 4% to 5% on that measure.

The release does not disclose sample size, fieldwork dates, respondent counts, question wording or margin of error. It therefore documents stated preference, not measured AI booking conversion, revenue share or transaction growth.

What does trust mean for AI booking?

SiteMinder Asia-Pacific market vice president Bradley Haines said: “As with all new technologies, there’s a trust stage that needs to pass before we see mass adoption of AI bookings.” He added that Australians retain abundant distrust of the technology.

His regional comparison put preference for AI assistants at 10.7% in China and 9.6% in Indonesia, against 5.0% in Singapore. Haines described Asia as dynamic, with AI readiness advancing at different speeds, and said hotels need better tools and data to identify where their demand originates.

Can AI create distribution upside?

Haines described hotel commerce as already moving toward “agentic,” meaning AI systems acting across booking steps, and cited global behavioral shifts. The release provides no booking-volume series or conversion data to test that assertion or quantify AI-driven distribution gains.

He said distribution technology, or tools that expand a hotel’s reach, and content can now make properties more visible through AI than ever before. The release offers no channel-lift evidence, commission changes (seller fees) or market-share results, so sellers face a claim about reach rather than proof of incremental bookings.

For travel sellers, Haines’s practical recommendation matters more than the terminology: use content and demand data to identify the source of each booking signal, then respond fast enough to convert visibility into revenue. Hotel operators still lack that speed, Haines said, without quantifying the gap.

The clearest revenue claim remains a projection. In a separate study, 65% of hoteliers globally expected immediate action on changes in travel demand to unlock at least 6% more revenue each year, Haines said. That finding records expectations, not audited revenue; no realized uplift appears in the release.

Even so, the claim shifts the commercial case toward demand-response systems and trusted AI content. It does not establish that hotels can count on a 6% gain, and sellers should test whether AI referrals deliver enough volume to justify new distribution spending.

Why users still book themselves?

Luke Chapman, a Tasmania-based marketing executive at a software company and founder of Japan-focused site Sake & Sakura, uses AI “all day, every day” at work but limits it in travel planning. He trusts it for initial research, checks every suggestion and makes bookings himself.

Chapman has encountered recommendations for places closed for renovations or permanently shut. He also objects when AI confidently gives wrong information, apologizes after a correction and then supplies the right answer; his concern is that repeated fluency can conceal errors until a traveler acts.

A Japanese museum once appeared open in an AI recommendation, but the official website showed it would close for renovations during Chapman’s trip. He said even a system that is right 95% of the time can eventually lead someone astray, potentially leaving a traveler stranded.

Payment marks Chapman’s firm boundary, especially for a large holiday purchase. He said: “I wouldn’t trust putting my credit card details into AI and letting it make the payment.” He wants to inspect the destination site, confirm it is the hotel’s official website or a booking platform he trusts, and enter card details himself.

That behavior favors established direct and online travel agency checkout paths (OTAs, the sites that sell hotel rooms) over fully delegated payment. It also gives hotels a practical distribution message: AI may help shoppers discover inventory, but accuracy, recognizable brands and controlled payment can determine whether those referrals convert.

What could AI change for sellers?

Chapman also believes generic prompts produce generic destinations. His AI recommendations for “Where should I go in Japan?” often offered places available through a quick Google search, crowding him toward popular sites; his better trips came from personal recommendations and conversations with locals.

For sellers, that limits AI’s near-term role in creating differentiated demand unless hotels supply specific, current content about opening status and local experiences. SiteMinder’s figures point to greater AI preference in some Asian markets, but Chapman’s experience shows that conversion still depends on trusted checkout paths, accurate information and rapid responses to shifting demand.