Artificial intelligence is becoming increasingly adept at mimicking human conversation, but a new analysis suggests it is not yet delivering truly personalized travel advice. According to a report by Skift Research, while AI models adapt their language to match a traveler’s specific profile, the actual destinations they suggest change very little.
The study, titled The Mystery of the Disappearing Destination, tested three major large language models: ChatGPT, Gemini, and Claude. Researchers used repeated prompts across various traveler personas, varying demographics, interests, travel companions, and budgets. The models successfully adjusted their tone and included relevant details, such as family-friendly activities for parents or low-cost flight options for budget-conscious users. However, the core recommendations remained remarkably static.
Concentration of Top Destinations
The findings indicate that AI recommendations are almost as concentrated as global international arrivals. The top five destinations accounted for 38% of all AI suggestions. This figure is close to the 40% share those same five locations hold in actual global inbound arrivals. At the top 10 level, AI recommendations reached 48%, compared to 56% for real-world arrivals.
This concentration suggests that AI still narrows the field in a way that traditional search engines do not. A standard search results page offers hundreds of links, while an AI answer typically returns a handful of options with attached reasoning. Despite this limitation, travelers appear to prefer the curated approach.
In a Skift Research and McKinsey survey of 1,015 U.S. travelers, 63% preferred a curated shortlist of 10 options. This was significantly higher than the 30% who preferred a 500-option catalog or the 7% who wanted a single editable AI itinerary.
Implications for Destination Marketing
The report concludes that destination marketing organizations should treat AI visibility as a distribution problem. Rather than relying solely on brand campaigns, marketers need to improve structured local content. They must also work with the operators and platforms that feed information into AI systems to ensure their destinations are included in the limited set of recommendations.
While AI personalizes the message, it does not yet personalize the recommendation. The technology is better at sounding like it knows the traveler than at actually discovering new places for them. This distinction is critical for the travel industry as it adapts to the growing influence of generative AI in the booking process.
Source: Skift
