Artificial intelligence (AI) is rapidly transforming the landscape of higher education, prompting universities to reconsider their pedagogical frameworks, support systems, and ethical guidelines. Despite the growing presence of AI tools in academic settings, institutional responses have often been reactive, focusing narrowly on academic integrity rather than on comprehensive integration strategies. This study investigates university teachers’ attitudes, usage patterns, and support needs related to AI in teaching at the University of Iceland, offering baseline data to inform future policy and faculty development initiatives.

In spring 2025, a 79-item survey and six open-ended questions were distributed to 2,003 university teachers, including sessional staff. A total of 339 responses were received (17% response rate; 33% among tenure-track faculty). Quantitative data were analyzed using descriptive statistics, while qualitative responses underwent thematic analysis using an inductive approach.

Findings reveal that while the majority of respondents hold positive attitudes toward AI, its integration into teaching remains limited. Most teachers use AI for supplementary tasks such as generating ideas for assignments (39%), developing course materials (35%), creating questions for exams (30%) and creating assessment rubrics (24%). Advanced applications—such as simulations, analytics, and chatbot training—were reported by a small minority. ChatGPT was the most widely used tool (62%), followed by Copilot (29%) and Gemini (14%). The highest subscription rate was for ChatGPT (12%), followed by Claude (3%), and both Scite and Copilot 365 at 2% each.

Key barriers to adoption include concerns about student misuse (89%), AI hallucinations (76%), ethical issues (71%), data privacy (69%) and bias in responses (68%). Teachers also cited a lack of institutional guidelines and information (62%), lack of support (49%) and limited personal knowledge (48%) as significant challenges. Open-ended responses emphasized the need for comprehensive training programs, discipline-specific, personalized support, university-funded access to AI tools, and clear policies on ethical and pedagogical use.

Teachers expressed cautious optimism about AI’s potential to enhance teaching efficiency and creativity but stressed the importance of preserving human elements in education. The study highlights a critical gap between interest and implementation, suggesting that universities must move beyond ad hoc responses and develop robust, interdisciplinary frameworks for AI integration.

This pilot study provides foundational insights into faculty experiences with AI at the University of Iceland and underscores the urgency of institutional action. Faculty development teams must prioritize hands-on training, ethical guidance, and infrastructure support to ensure educators are equipped to navigate the evolving educational landscape. Future research should explore variations in attitudes across disciplines, age groups, and employment status, and monitor longitudinal changes as AI tools become more embedded in academic practice.

Keywords: Artificial intelligence, higher education, university teachers, teacher perceptions, AI in teaching, faculty attitudes, educational technology, AI support needs, AI policy, teacher development, professional learning, ChatGPT, Copilot, ethical concerns, data security, content quality, AI tools in academia, instructional support, academic innovation, survey research.