Which? Travel Investigation Exposes Flaws in Tripadvisor’s AI
An investigation by Which? Travel has revealed that Tripadvisor’s artificial intelligence system misleads travelers by suppressing negative reviews. Focusing on the five-star-rated Riu Palace hotel in Cabo Verde, the study found that the AI overlooked numerous complaints about food poisoning, poor hygiene, and inadequate cleaning standards. Guests reported raw food, flies, and birds in their meals. Which? Travel identified over a hundred such reviews for this property.
Riu Hotels & Resorts insists it follows the strictest international health and hygiene standards. However, according to the investigation, seven deaths have been recorded at the hotel since early 2023. As a result, a class-action lawsuit has been filed against the hotel, representing at least 412 vacationers.
Misleading Information on Tripadvisor
Tripadvisor’s platform rates the hotel’s cleanliness as impeccable. Yet, another AI tool on the site—a chatbot named Ollie—failed to warn users about potential risks tied to poor hygiene. When asked about the likelihood of food poisoning, Ollie responded that it was ‘quite unlikely.’
Hotel guests expressed their frustration, noting: ‘The food was terrible, tasteless, dangerous, and inedible’ — a visitor’s review.
A Tripadvisor spokesperson stated that the company ‘prioritizes transparency and impartiality.’ They added that summaries reflect a broad range of both positive and negative community reviews. According to the spokesperson, Ollie selects reviews based on detail and relevance. They also noted that the Ollie product is still in development.
The investigation highlights potential shortcomings in the use of AI within the travel industry, which could have serious consequences for traveler safety.
This case underscores the need for transparency and accuracy in information provided by online travel platforms. Overlooking negative reviews can lead to severe health and safety risks for vacationers who rely on such data when choosing destinations. Consequently, companies must reevaluate how they deploy artificial intelligence to deliver more reliable and secure recommendations to users.