Grand Challenges for Research on Privacy Documents (Dagstuhl Seminar 25021): Datasets, Automation, and Analysis Methods

Story, Peter; Ghanavati, Sepideh; Hosseini, Henry; Mitrovic Jelena; Samples, Tim; Wagner, Isabel; Zhao, Tianyang

Abstract

This report documents the program and the outcomes of Dagstuhl Seminar 25021 "Grand Challenges for Research on Privacy Documents" held in January 2025. This Dagstuhl Seminar gathered an interdisciplinary group of researchers from privacy, natural language processing, human-computer interaction, public policy, and law to identify and characterize key challenges to research on privacy documents, such as privacy policies, terms of use, cookie policies, and other texts about data practices.
Seminar participants worked together to identify and characterize key challenges in privacy document research with the goal of producing a research roadmap for tackling these challenges. Through a series of perspectives talks and panel discussions, participants exchanged experiences in working with privacy documents in research and learned about associated challenges, as well as interdisciplinary intersections and policy considerations. Through deeper engagement in working groups, participants deeply explored research challenges and research directions across five interconnected topics: (1) formats and standardization; (2) datasets, automation, and analysis methods; (3) usable and useful notice and consent; (4) consumer privacy beyond notice and choice; and (5) cross-stakeholder engagement.

Keywords
Human-Computer Interaction; Machine Learning; Natural Language Processing; Privacy Policy; Public Policy