AI summaries of Tripadvisor hotel reviews understate complaints, investigation finds
AI Summaries of Tripadvisor Hotel Reviews Understate Complaints, Investigation Finds
AI summaries of Tripadvisor hotel reviews - A recent investigation has revealed that Tripadvisor’s new AI-generated review summaries for hotels do not adequately highlight serious complaints, such as food poisoning, sexual harassment, and hygiene issues, according to consumer advocacy group Which?. The findings suggest that the automated system may present an overly positive image of accommodations, potentially misleading travelers about their true quality and safety.
The AI Summary System and Its Limitations
Which? conducted tests on Tripadvisor’s AI tool, which aggregates user reviews into concise summaries displayed on the hotel booking platform. These summaries aim to provide a quick overview of recent feedback, yet the investigation found that they often omit critical details. For instance, the AI failed to emphasize recurring issues of poor hygiene or food safety, even when multiple guests raised similar concerns.
During its analysis, Which? noted that the AI summaries prioritize positive language, such as phrases like “popular with many travelers” or “diverse restaurants” that “earn rave reviews.” These descriptions can create an impression of excellence, even when serious problems are reported. In one example, the AI summary for the Riu Palace Santa Maria, a five-star all-inclusive resort in Cape Verde, highlighted “spotless cleanliness” and “friendly service,” despite dozens of complaints from recent guests.
Concrete Examples of Missed Complaints
One specific case involved the Riu Palace Santa Maria, where reviewers described a range of hygiene failures. Guests reported serving raw chicken, the presence of flies in buffet food, and even “dead little roasted mice by the sitting area” during their stay. These issues were mentioned in over 100 reviews, including 14 that stated at least one person in the party had fallen seriously ill due to food poisoning. However, the AI summary did not reflect these concerns, instead focusing on favorable aspects.
Similarly, the AI summary for a Dominican Republic hotel praised its “abundant amenities,” but barely touched on “inconsistent cleanliness” and “maintenance issues” cited by several travelers. In another instance, a hotel in Turkey was highlighted for its “friendly service,” despite reports of sexual harassment by male staff, including inappropriate jokes, gestures, and repeated requests for social media connections. Reviewers described feeling “unsafe” during their stays, but the summary only mentioned “lapses in service noticed by a few.”
Which? emphasized that the AI tool’s design may lead to an incomplete portrayal of a hotel’s performance. “The summaries are generated by selecting the most relevant reviews, but they do not always capture the full scope of complaints,” a spokesperson for Which? stated. This could result in travelers making decisions based on an inaccurate summary, potentially overlooking critical issues that affect their experience.
Responses from Tripadvisor and RIU Hotels
“The health and safety of our guests is always our main priority,” said a representative from RIU Hotels & Resorts, responding to the findings. They added: “Let us assure that we operate with the highest standards of professionalism and service, placing hygienic-sanitary safety as our top priority. Our hotels in Cape Verde follow strict international health and hygiene standards, certified by external prestigious consultancy firms.”
Tripadvisor defended its AI summaries, stating they are designed to uphold “integrity and transparency.” A spokesperson explained that the summaries are based on high volumes of user-generated content and are not meant to replace individual reviews. “Users can easily click to see the traveler quotes behind each review element or access all reviews for that listing, eliminating any need to blindly trust AI-generated content,” the representative said.
Additionally, Tripadvisor highlighted its safeguards, claiming that the AI system automatically suppresses summaries for listings with warnings about serious safety incidents. These include cases of death, drugging, or sexual assault, ensuring such content is prominently displayed. However, the investigation found that even with these measures, critical complaints were still underrepresented in the summaries.
Expanding the Scope of the Investigation
Which? also tested Tripadvisor’s AI chatbot, which was found to fail in warning users about poor hygiene conditions at certain hotels. The chatbot, according to Tripadvisor, draws from a selection of reviews based on detail and recency, matching them by language and context. “We are actively looking into the examples where reviews did not match the intended purpose,” the spokesperson said, acknowledging the need for further refinement.
The investigation’s findings raise questions about the reliability of AI summaries in capturing the full range of traveler experiences. While the tool is intended to streamline information for users, its current design may prioritize positive feedback over negative, creating a skewed perception of hotel quality. For example, the AI’s failure to mention hygiene lapses at the Riu Palace Santa Maria or sexual harassment incidents in Turkey suggests that it may not effectively identify patterns or recurring issues.
Impact on Travelers and the Hospitality Industry
Consumers who rely on AI summaries may unknowingly choose hotels with significant drawbacks. This could lead to dissatisfaction, as guests might expect a certain standard based on the summary but encounter issues firsthand. The problem is compounded by the fact that negative reviews often require more effort to find, as the AI tool does not always highlight them prominently.
“The summaries are supposed to give a balanced view, but in reality, they seem to favor the positive over the negative,” said one Which? analyst. The group argues that the AI system needs improvements to ensure it accurately reflects both praise and criticism. This would help travelers make more informed decisions and hold hotels accountable for their shortcomings.
Tripadvisor’s response underscores its commitment to transparency, but the investigation highlights areas for enhancement. The spokesperson noted that the AI system is continuously monitored and refined, with efforts to capture all types of feedback. However, the examples provided by Which? suggest that the current process may still miss important details, especially when complaints are not immediately obvious or are spread across multiple reviews.
Calls for Greater Accountability and Transparency
Which? has called for more accountability from Tripadvisor, urging the platform to ensure its AI summaries do not downplay serious issues. The group’s findings align with broader concerns about the role of AI in shaping consumer perceptions. As more travelers use automated tools to assess their options, the accuracy of these summaries becomes increasingly critical.
“AI can be a powerful tool, but it must be used responsibly,” said a Which? representative. “If the summaries fail to capture the full range of complaints, they risk misleading users and undermining trust in the platform.” The investigation serves as a reminder that while AI can enhance user experience, it also needs to be carefully designed to reflect all aspects of a hotel’s performance.
In conclusion, the study reveals a gap between the positive language used in AI summaries and the actual experiences reported by travelers. By underrepresenting serious complaints, the system may not provide the comprehensive insight needed for informed decision-making. As the hospitality industry continues to evolve, it is essential that platforms like Tripadvisor adapt their tools to ensure they accurately reflect the quality and safety of the accommodations they promote.