Ethical AI in tourism analytics: is machine unlearning the missing puzzle?

This research proposes Machine Unlearning (MU) as a technical counterpart to legal data-erasure rights. It explores its relevance for tourism analytics, where sensitive traveller data and transnational laws make privacy a critical issue. Focussing on its conceptual contribution, it aims to (a) simplify MU approaches and demonstrate their applicability in tourism, (b) identify regulatory gaps that leave data traces inside trained models even after deletion, and (c) illustrate, via a controlled simulation, how removing a user through MU can change the recommendation outputs. The key contribution is positioning MU as a bridge between legal erasure rights and hidden data retention in tourism AI pipelines. The letter further identifies implications for tourism analytics including accountable recommender design and cross-jurisdictional data governance. It concludes by outlining the practical implications of MU and future directions for cross-border implementation, governance innovation, trust building, and research in the data-driven tourism sector. The study is conceptual in nature and is complemented by a small, non-generalisable illustrative simulation.