User Trust and AI-Enhanced Transparency: A Study on Cookie Banner UI Design
2025 (English)In: HCI for Cybersecurity, Privacy and Trust / [ed] Moallem, A., Springer, 2025, p. 128-147Conference paper, Published paper (Other academic)
Abstract [en]
Since the introduction of the General Data Protection Regulation (GDPR) in 2016, the design of cookie banners and the use of cookies themselves have to follow guidelines. Consent has to be given freely, informed, and all information has to be specific and unambiguous. However, a survey of almost 30.000 websites in 2022 showed that 94.7% of them did not fully comply with these set guidelines. Dark patterns are deceptive design practices that often lead to user confusion and reduced autonomy. Artificial Intelligence (AI) tools, such as plug ins, could help users navigate and mitigate these patterns. However, building user trust in such tools remains a challenge. This user study, conducted with a developed prototype, aims to gain insights into user perspectives on trust, behavior, and consent decisions while also exploring opinions on AI assistance for managing cookie consent. The prototype features different cookie banner designs, including a dark-pattern UI, a GDPR-compliant UI, and an AI-assisted approach. The evaluation consists of A/B testing of these UI alternatives, combined with a survey. Statistical analysis and sentiment/thematic analysis of data collected from 362 participants show the importance of transparency for building trust when designing cookie banners. Furthermore, our results show that while GDPR-compliant banners are more time-consuming to navigate, they empower users to make more informed consent decisions. Participants demonstrated a learning curve, becoming more efficient over time. Additionally, qualitative feedback revealed skepticism toward AI-assisted approaches when the AI actions were opaque (black-boxed), reinforcing the importance of transparency for trust.
Place, publisher, year, edition, pages
Springer, 2025. p. 128-147
Series
Lecture Notes in Computer Science, ISSN 0302-9743, E-ISSN 1611-3349 ; LNCS, volume 15815
Keywords [en]
Transparency, cookie banner, GDPR compliance, UI design, privacy, AI support, dark patterns
National Category
Human Computer Interaction
Research subject
Information Systems
Identifiers
URN: urn:nbn:se:kau:diva-105852DOI: 10.1007/978-3-031-92840-6_8ISI: 001572075600008Scopus ID: 2-s2.0-105008998140ISBN: 978-3-031-92839-0 (print)ISBN: 978-3-031-92840-6 (electronic)OAI: oai:DiVA.org:kau-105852DiVA, id: diva2:1977640
Conference
7th International Conference, HCI-CPT, Gothenburg, Sweden, June 22–27, 2025.
2025-06-262025-06-262026-02-12Bibliographically approved