Safety Warnings Scrubbed from AI Summaries
The British consumer advocacy group Which? has accused Tripadvisor of failing to accurately represent guest experiences in its AI-generated hotel review summaries. According to the organization, these automated summaries often omit critical warnings regarding food poisoning, poor hygiene standards, and reports of harassment, potentially misleading travelers about safety conditions.
The Riu Palace Santa Maria Disconnect
The investigation focused on the effectiveness of AI tools in synthesizing large volumes of guest feedback. A primary example cited by the group is the five-star Riu Palace Santa Maria on the Cape Verde islands. While the Tripadvisor AI summary characterized the hotel as “popular” with “spotless” cleanliness, Which? found that individual user reviews contained numerous reports of poor hygiene, bird and insect presence in buffet areas, and instances of raw food.
In a March 2024 audit, Which? identified 102 distinct mentions of food poisoning associated with the property. The hotel chain, Riu, maintains that the health and safety of its guests remain its highest priority, stating that its facilities comply with rigorous international hygiene standards verified by external auditing firms. Reports indicate that a class-action lawsuit involving at least 412 travelers who allegedly fell ill at the resort is currently underway.
Chatbot Responses Under Scrutiny
Beyond static summaries, Tripadvisor utilizes an AI chatbot named “Ollie” to answer user queries. Which? reported that when asked specifically about the risk of food poisoning at the Riu Palace, the chatbot suggested such occurrences were unlikely, citing the hotel’s reputation for high hygiene standards.
In another instance involving a hotel in Antalya, Turkey, the consumer group found that the AI summary categorized service as “friendly,” despite multiple guest reports detailing instances of sexual harassment by staff members, including reports of employees following guests to solicit their social media profiles.
Google’s Contrasting AI Performance
The investigation contrasted Tripadvisor’s results with AI-generated summaries provided by Google. In the case of the same hotels, Google’s summaries included references to potential health risks and reports of sanitary issues, such as the presence of birds in dining areas. Rory Boland, travel editor at Which?, characterized the omission of safety-critical information as “unacceptable” and potentially dangerous for travelers.
Tripadvisor Defends Automated Tools
Tripadvisor has contested the findings of the Which? report. The company stated that its AI summaries are designed to reflect a balance of both positive and negative feedback and are intended to supplement, not replace, individual reviews. The platform noted that users retain the ability to access original reviews with a single click.
Furthermore, the company clarified that its systems are programmed to automatically suppress AI summaries when reviews contain warnings of severe incidents, such as fatalities, sexual assault, or drugging. Tripadvisor maintains that the accusation that it suppresses content to the detriment of traveler safety is unfounded.
Vetting Hotels Beyond the Algorithm
Which? advises travelers to exercise caution when relying on automated summaries. The organization recommends that users skip AI-generated overviews entirely and instead prioritize reading individual reviews, particularly those with one-star ratings, to gain a more accurate understanding of a hotel’s safety and service standards. Consulting multiple review platforms remains a recommended practice for verifying property conditions before booking.
Keep reading