Key Points
- AI’s biggest influence is upstream: it shapes which destinations enter a traveler’s shortlist, and query phrasing alone can make a destination appear first or vanish entirely—so destinations must optimize for how travelers actually ask.
- AI favors destinations with easy visa access, short flight times, strong safety records and abundant Chinese-language content, creating a compounding feedback loop that strengthens already-popular ‘winner’ destinations.
- AI currently influences discovery and consideration but not the transaction—44% of Chinese travelers use AI for planning, yet they still rely on OTAs to compare and book, keeping AI complementary rather than a replacement.
Summary
A study by Create Consulting China, which tested 240 travel queries across five major Chinese AI platforms (Baidu’s ERNIE Bot, ByteDance’s Doubao, Alibaba’s Tongyi Qianwen, DeepSeek and Tencent Yuanbao), finds that AI’s most consequential influence on travel happens early—at the “consideration set” stage—by determining which destinations even make it into a traveler’s shortlist. The research shows that how a question is phrased dramatically alters recommendations: broad, open-ended queries cluster around a handful of safe, familiar, visa-friendly and nearby destinations (Singapore, Thailand, Japan), while continent-specific queries surface places like South Africa and Mexico that never appeared in open-ended results at all. AI tends to be more conservative than travelers themselves, weighting visa access, flight distance, safety and—critically—the availability of Chinese-language content, which creates a compounding feedback loop favoring already-popular destinations. Notably, AI mirrors genuine human attention surges (as seen with Cape Verde during the 2026 World Cup) rather than driving them. For now, travelers delegate the consideration set to AI but still turn to OTAs to compare and complete the actual transaction.
