Blab AI
Tripadvisors AI Summaries Mask Serious Hotel Safety Issues, Says UK Consumer Group
A July 2, 2026 investigation by the UK consumer watchdog Which? found that Tripadvisor’s new AI‑generated review summaries and its Ollie chatbot are downplaying or omitting serious safety, hygiene and harassment complaints at several hotels.
The investigation focused on four properties that have attracted large numbers of negative reviews. At the five‑star, all‑inclusive Riu Palace Santa Maria in Cape Verde, guests reported food poisoning, raw chicken, flies and mice on buffet tables, and a group legal action involving more than 400 ill travelers and seven deaths since 2023. The AI summary, however, described the resort as “spotless” and praised its dining, while Ollie said food poisoning was “quite unlikely” and highlighted high hygiene standards.
Similar patterns emerged at other resorts. The Garza Blanca resort in Cancun, where a wedding party and other guests fell violently ill, was described by the AI as having “immaculate cleanliness.” The Occidental Caribe in the Dominican Republic, where reviews cited a lack of running water, rooms smelling of raw sewage and widespread mold, was summarized as offering “abundant amenities” and reducing structural problems to “maintenance issues.” At the Kaia Coracesium on Turkey’s Antalya coast, multiple female guests reported sexual harassment by staff, including staff following guests to their rooms and demanding social media details. The AI summary called the service “friendly” and framed the incidents as “lapses in service.”
Tripadvisor said its algorithms analyze reviews from the past year to highlight the most common themes and that the summaries are meant to be snapshots, not replacements for reading full reviews. The company also noted that travelers should not “blindly trust” the AI.
Which? editor Rory Boland criticized the placement of the summaries at the top of hotel pages, arguing that the prominence encourages users to rely on them. He said that “by burying life‑threatening risks under fluffy, automated praise, the platform is putting profits and convenience ahead of human lives.”
The Guardian, Euronews, and other outlets reported on the investigation, noting that Google’s AI, in contrast, flags outbreaks of illness and potential disease risks at the same properties. Which? said that until AI models are trained to prioritize safety over marketing language, travelers should read the full reviews, especially the one‑star accounts.
The Riu Palace Santa Maria case is part of a broader legal dispute. A group action filed in 2023 includes more than 400 claimants seeking compensation for illness and, in some cases, death. The resort’s management has not publicly responded to the Which? findings.
Tripadvisor’s AI tool, launched in June 2026, was promoted as a way to “make travel planning smoother and more insightful.” The company’s statement emphasized that the summaries are designed to help users quickly assess common themes.
The Which? investigation highlights the risk of AI hallucinations or omissions in high‑stakes contexts. When an algorithm filters user‑generated content for brevity, it can inadvertently suppress critical safety information. The findings have prompted calls for clearer labeling of AI‑generated summaries and for users to verify information through multiple sources.
At present, no regulatory action has been announced. TripAdvisor has not indicated plans to modify its AI summarization approach. Travelers are advised to consult the full range of reviews and to remain cautious of AI‑generated overviews that may omit serious concerns.
The investigation underscores the need for transparency in AI‑driven content curation, especially in industries where safety and health are paramount. Until such measures are in place, consumers must rely on detailed, unfiltered reviews to make informed travel decisions.
The investigation focused on four properties that have attracted large numbers of negative reviews. At the five‑star, all‑inclusive Riu Palace Santa Maria in Cape Verde, guests reported food poisoning, raw chicken, flies and mice on buffet tables, and a group legal action involving more than 400 ill travelers and seven deaths since 2023. The AI summary, however, described the resort as “spotless” and praised its dining, while Ollie said food poisoning was “quite unlikely” and highlighted high hygiene standards.
Similar patterns emerged at other resorts. The Garza Blanca resort in Cancun, where a wedding party and other guests fell violently ill, was described by the AI as having “immaculate cleanliness.” The Occidental Caribe in the Dominican Republic, where reviews cited a lack of running water, rooms smelling of raw sewage and widespread mold, was summarized as offering “abundant amenities” and reducing structural problems to “maintenance issues.” At the Kaia Coracesium on Turkey’s Antalya coast, multiple female guests reported sexual harassment by staff, including staff following guests to their rooms and demanding social media details. The AI summary called the service “friendly” and framed the incidents as “lapses in service.”
Tripadvisor said its algorithms analyze reviews from the past year to highlight the most common themes and that the summaries are meant to be snapshots, not replacements for reading full reviews. The company also noted that travelers should not “blindly trust” the AI.
Which? editor Rory Boland criticized the placement of the summaries at the top of hotel pages, arguing that the prominence encourages users to rely on them. He said that “by burying life‑threatening risks under fluffy, automated praise, the platform is putting profits and convenience ahead of human lives.”
The Guardian, Euronews, and other outlets reported on the investigation, noting that Google’s AI, in contrast, flags outbreaks of illness and potential disease risks at the same properties. Which? said that until AI models are trained to prioritize safety over marketing language, travelers should read the full reviews, especially the one‑star accounts.
The Riu Palace Santa Maria case is part of a broader legal dispute. A group action filed in 2023 includes more than 400 claimants seeking compensation for illness and, in some cases, death. The resort’s management has not publicly responded to the Which? findings.
Tripadvisor’s AI tool, launched in June 2026, was promoted as a way to “make travel planning smoother and more insightful.” The company’s statement emphasized that the summaries are designed to help users quickly assess common themes.
The Which? investigation highlights the risk of AI hallucinations or omissions in high‑stakes contexts. When an algorithm filters user‑generated content for brevity, it can inadvertently suppress critical safety information. The findings have prompted calls for clearer labeling of AI‑generated summaries and for users to verify information through multiple sources.
At present, no regulatory action has been announced. TripAdvisor has not indicated plans to modify its AI summarization approach. Travelers are advised to consult the full range of reviews and to remain cautious of AI‑generated overviews that may omit serious concerns.
The investigation underscores the need for transparency in AI‑driven content curation, especially in industries where safety and health are paramount. Until such measures are in place, consumers must rely on detailed, unfiltered reviews to make informed travel decisions.
