ИИ-резюме Tripadvisor назвал отель под судом «безупречным», скрыв жалобы на домогательства

Tripadvisor's AI-summary called a hotel under lawsuit "spotless," hiding harassment complaints

On July 2, 2026, the British consumer protection organization Which? published an investigation: the AI review summarization feature on Tripadvisor systematically conceals serious guest complaints — including sexual harassment and mass food poisoning incidents.

What the algorithm hides

Which? researchers examined AI summaries for several hotels and documented a systemic pattern: the algorithm does not simply downplay negativity, but in some cases generates a characterization directly contradicting the actual guest experience.

• A hotel facing a lawsuit over mass food poisoning was described by the algorithm as "spotless"
• A resort where guests extensively complained about sexual harassment by staff received a "friendly service" rating in the AI summary
• In other cases, the algorithm omitted complaints about mold smell and lack of centralized water supply
• Tripadvisor deployed automatic summarization to aggregate millions of reviews worldwide

The gap between the AI summary and actual reviews here is not a technical error like an inaccurate star count. It is a qualitative inversion: "spotless" and "lawsuit over mass poisoning" are mutually exclusive characteristics of the same establishment. The algorithm did not err in details — it reached a false conclusion.

Why algorithmic neutrality is dangerous

Tourists increasingly make booking decisions based on the AI summary at the top of the page rather than reviewing the entire corpus of comments. The convenience is obvious: no need to scroll through hundreds of comments spanning years. But if the algorithm systematically downplays negatives, users lose information critical to their own safety.

An algorithm trained on millions of reviews is likely optimized for "balance" — without explicit bias toward either criticism or praise. Such neutrality is acceptable when disagreements involve taste preferences: some guests like breakfast while others do not. But when complaints concern health and safety — food poisoning, harassment, unsanitary conditions — algorithmic neutrality becomes misinformation.

"Tourists should be able to trust what they read," states

Which? in its investigation conclusion.

A separate issue is platform accountability. When a human editor publishes misleading content, responsibility is clear. When an AI deployed by the platform itself does so, the legal landscape becomes murky. The Which? investigation implicitly raises this question: does Tripadvisor bear responsibility for an AI summary that potentially exposes users to risk?

What this means

The Which? investigation is another signal for the industry: automatic summarization of user-generated content requires special handling protocols for safety-related complaints. Reviews of poisoning and harassment are not simply "negative opinions" in the general category, but warnings that the algorithm must surface prominently, not average out. For Tripadvisor, this is a matter not only of reputation but of trust in the platform as a tool for genuine consumer decisions.

Frequently asked questions

How does Tripadvisor's AI summarization work?

The algorithm analyzes thousands of user reviews for a hotel and generates a brief text summary displayed at the top of the property page. The feature was designed to help users who lack time to read the entire corpus of comments. According to the Which? investigation, it is precisely this summary that conceals critically important complaints.

Who conducted the investigation?

Which? is an independent British nonprofit consumer protection organization, founded in 1957. It regularly tests digital products and publishes warnings about consumer risks, including in the AI services sector.