How is teachers’ use of artificial intelligence associated with their time spent on routine tasks, and how does this vary across gender and professional experience?

Recent research shows that an increasing share of teachers’ working time in the classroom since 2018 has been devoted to administrative activities, while a relatively smaller share of time is spent on actual teaching and learning (OECD TALIS 2024: 24). A substantial part of teachers’ work therefore consists of routine tasks such as administrative activities, planning, and communication. At the same time, a high workload of routine tasks may contribute to work-related stress, and approximately 35% of Danish teachers report that excessive preparation and related tasks constitute a source of stress in their work (OECD TALIS 2024). At the same time, developments in artificial intelligence (AI) within the education sector are progressing rapidly. The TALIS report itself emphasizes that “developments in this area are moving extremely quickly, and one can expect things to look different by autumn 2025” (OECD TALIS 2024: 47). This makes it particularly relevant to investigate AI use among teachers at the present moment, when the technology is being integrated into practice. In Denmark, for example, approximately 90% of schools already use the digital platform Skoletube, which includes generative AI functionalities accessible to both teachers and students. Teachers constitute a particularly relevant occupational group for studying the potential implications of AI. A significant share of teachers’ working time is devoted to routine tasks that may potentially be supported by AI, including administrative work, planning, and written communication. At the same time, teachers represent a large public-sector workforce, meaning that even relatively small changes in work practices could have substantial aggregate effects. In 2024, approximately 36% of teachers reported having used AI in their work (OECD TALIS 2024). Given the rapid diffusion of AI tools, it is reasonable to expect that this share has increased. Against this background, this project examines the relationship between teachers’ use of artificial intelligence and their time allocation to routine tasks. More specifically, the study investigates whether the introduction and use of AI are associated with deployment effects, understood as a reduction in the time teachers spend on routine tasks when parts of these tasks are supported or automated by AI. In addition, the study examines how AI adoption varies across gender and years of professional experience. The overall aim is to contribute new empirical knowledge on whether the adoption of AI may be associated with changes in teachers’ time use and work practices. As contextual background, the analysis will also include a descriptive comparison of teachers’ time spent on routine tasks with figures reported in TALIS 2024. While the data sources are not directly comparable - since the TALIS survey is based on a stratified probability sample whereas our data rely on a different sampling approach- the comparison may nevertheless provide a tentative indication of potential developments in teachers’ time use between 2024 and 2026. This comparison is therefore intended to offer descriptive context rather than a formal hypothesis test. Understanding these potential deployment effects is important because even small reductions in time spent on routine tasks could have substantial aggregate implications in the education sector. Time saved through AI-supported tasks may either reduce overall labor requirements or allow teachers to reallocate time toward more complex pedagogical activities.