The Transparency Issue of a Black Box: Exploratory research study on AI communication in higher education to explore students’ perceptions of transparent AI use and disclosure practices
2026 (English) Independent thesis Advanced level (degree of Master (One Year)), 10 credits / 15 HE credits
Student thesis
Sustainable development Sustainable Development Alternative title
The Transparency Issue of a Black Box : Exploratory research study on AI communication in higher education to explore students’ perceptions of transparent AI use and disclosure practices (English)
Abstract [en]
The increasing role of generative artificial intelligence (GenAI) in higher education has prompted universities to introduce guidelines promoting responsible use and transparent disclosure of AI-assisted academic work. However, it remains unclear whether and to what extent such institutional guidelines provide students with practical and pedagogical guidance. The study addresses this policy-practice gap by examining students’ perspectives on existing AI guidelines and their pedagogical enactment. Using a mixed-method approach, the study followed a three-stage process comprising a pre-intervention survey, a pedagogical intervention, and a post-intervention survey. Thirty-one students from diverse study backgrounds, enrolled in two German courses at a Finnish university, participated in the study and evaluated an approach designed to support the transparent disclosure of GenAI. The analysis adopts the three dimensions of Chan’s (2023a) AI Ecological Education Policy Framework: Pedagogical, Operational, and Ethical. The aim is to examine students’ perceptions of AI in higher education, their compliance with transparency requirements, and actual disclosure practices. The findings unravelled three distinct student positioning patterns: (1) rejection of AI use and distrust toward AI systems, (2) conditional acceptance accompanied by ethical and operational uncertainty, and (3) pragmatic and confident AI use. While the quantitative results indicated a general awareness of institutional AI guidelines, qualitative responses revealed a lack of clarity, consistency, and practical examples regarding the use and disclosure of GenAI in academic contexts. This lack of guidance contributed to uncertainty, concerns about unintentionally violating academic expectations, as well as distrust regarding the legitimacy of AI use. Overall, the study demonstrates that the effectiveness of institutional AI policies depends not only on their existence but also on their pedagogical enactment through assessment design and reciprocal communication between teachers and students.
Place, publisher, year, edition, pages 2026. , p. 56
Keywords [en]
transparency, AI guidelines, academic integrity, black box, AI communication
National Category
Educational Sciences
Identifiers URN: urn:nbn:se:hj:diva-73754 OAI: oai:DiVA.org:hj-73754 DiVA, id: diva2:2099592
Subject / course HLK, Education
Supervisors
Examiners
2026-09-162026-09-082026-09-16 Bibliographically approved