Universities and Colleges Must Prepare for New EU AI Regulations
Article Content
The rapid integration of artificial intelligence (AI) in higher education and research is prompting universities and colleges in Norway to ready themselves for new regulatory requirements. An extensive EU AI regulation is set to be incorporated into Norwegian law, imposing new obligations on institutions that utilize AI in teaching, assessment, research, and administration.
According to the 2025 Study Barometer, over 70% of students frequently or occasionally use AI in their academic work, underscoring AI’s growing importance in the sector. However, not all institutions have fully assessed the risks and responsibilities associated with AI use. Tommy Tranvik, senior advisor at the Directorate for Higher Education and Competence (HK-dir), emphasizes that preparations must begin immediately rather than wait for the regulation to take effect.
Tranvik outlines five key actions for educational leaders: mapping all AI applications across teaching, research, and administration; building competence on regulatory requirements; clarifying institutional roles as either AI providers or operators; assessing the risk categories of AI systems; and establishing clear roles, procedures, and internal controls. These steps are critical for ensuring compliance with obligations related to risk assessment, documentation, privacy, and human oversight.
Høgskolen i Østfold (HiØ) began its AI readiness efforts two years ago and has since expanded them into a three-year project focused on competence development and new AI services. The institution is introducing principles for AI use in examinations, inspired by the University of Tromsø, to foster responsible and transparent AI practices. Marianne Bjerkman, head of IT development at HiØ, highlights the importance of distinguishing between using generative AI as a learning tool and as a means to simplify tasks, noting that this difference is not always clear.
As AI technologies become more embedded in higher education, institutions must address governance, accountability, and trust to meet the demands of evolving regulations and ensure ethical AI integration.
Read more: kode24.no