Tripadvisor’s AI Is Telling Vacationers That Gruesome Horror Grottos Are Wonderful Holiday Retreats
TL;DR
A Which? investigation says Tripadvisor’s AI review summaries are softening serious hotel complaints instead of clearly warning travelers. Examples include a Cape Verde hotel described as clean despite lawsuits over mass food poisoning, and a Turkey resort where harassment complaints were reduced to vague service lapses. Tripadvisor says it suppresses AI summaries for serious safety incidents and is reviewing the cases. That answer sounds polished, but does not settle the reliability issue.
Nauti's Take
For review summaries, the first check is loss control: which severe individual reports disappear in the compressed version? Teams should compare AI output against the worst-case reviews and explicitly prioritize safety, hygiene, or fraud signals instead of treating a balanced tone as a quality marker.
Briefingshow
Reviews are supposed to surface risk, not polish it away. When a platform puts AI between travelers and guest reports, it changes more than length: rare but severe warnings can disappear inside a friendly average. That may be convenient for hotels and platforms, but it can make decisions worse for users.
Sources