Regulating Generative Artificial Intelligence Travel Agents to Protect Tourist Consumers from Algorithmic Misinformation

Abstract

Useful advice together with plausible inaccuracies – the protection offered by disclosure and consumer choice are in question with generative AI travel agents. This research combines a secondary analysis of a public survey published in 2023 and a targeted analysis of the regulation by the European Union and United Kingdom. The data set has 900 consumers with 886 complete cases for the adjusted models. Futurismo asked 224 people to indicate their willingness to make use of AI for travel planning in the future, with a mean perceived score of at least 4 on a five-point scale, which is considered to represent moderate or high likelihood; 58 of these said they would use AI for travel planning in the future (25.9% [95% confidence interval 20.6–32.0%]). The standardized risk–intention association was shifted from 0.181 to −0.049, with experience and related perceptions for acceptance being controlled. Such observational estimates do not provide misinformation detection or the effectiveness of warnings. Legal evaluation of the origin of the AI has been separated from the accuracy and commercial presentation of the travel advice. The paper recommends proportional verification of consequential claims, confirmation prior to the making of financial commitments, correction of these claims and the existence of a clearly-defined trader who can handle complaints. Factual reliability and recoverability should be considered along with disclosure compliance as a component of tourism authorities' evaluation of whether safeguards are factual and how they can be recovered, before they can be implemented. Factual testing should be carried out before the implementation of safeguards, along with factual reliability and recoverability.