Within doctoral education, generative AI raises specific questions, including methodology, transparency, responsibility and ethics. According to Deputy Vice-Chancellor Karin Axelsson, open discussions are essential, both within research environments and in the relationship between doctoral student and supervisor.
The seminar, Finding the way, marked the start of a university-wide dialogue on how generative AI can be used in a responsible, transparent and purposeful way. The seminar series is part of LiU鈥檚 action plan for AI proofed education, which includes both skills development and pedagogical development at all levels. A central concept in this context is AI literacy, the ability to understand, use and critically engage with AI tools in different contexts.
鈥淲e need to strengthen AI literacy among teachers, doctoral students and supervisors. At the same time, we must review how education and assessment are designed, so that students gain the right knowledge for a future working life,鈥� says Karin Axelsson.
Interest was high when doctoral students, supervisors and university leadership gathered to discuss both opportunities and challenges. Katherine Harrison organises the seminar series.
鈥淢any people are calling for guidelines and a way forward. This seminar is a response to those conversations,鈥� she says.
Policy built on trust
The seminar featured Eva 脜kesson, professor of chemical physics and senior advisor, and Rachel Forsyth, educational developer, both from Lund University. They shared their work on developing a university-wide policy for generative AI.
鈥淭here is a strong demand for clear rules, but this is a complex issue. It cannot be solved through detailed regulations alone,鈥� says Eva 脜kesson.
Instead, Lund University chose to develop a policy based on principles, with trust as a key starting point.
鈥淲e need to begin by saying: we trust you. If students and staff feel they have to hide how they use AI, then we have created the wrong kind of culture.鈥�
The policy also emphasises transparency and responsibility. Anyone using generative AI is always responsible for the outcome, regardless of whether it concerns text, code or other types of content.
As part of the policy work, Lund University also interviewed alumni, providing an up-to-date picture of how generative AI is used across different sectors and which skills are in demand in the labour market. The use varied widely between workplaces: some had strict rules and their own tools, while others had no guidelines at all. Many described learning to use AI through their own exploration.
Focus on dialogue 鈥� not control
A recurring theme throughout the seminar was the importance of dialogue rather than control. Rachel Forsyth warned against turning teachers into AI enforcers.
鈥淲e do not believe it is possible to reliably determine whether something has been created using AI. And we risk damaging the trust between teachers and students if we become stuck in that way of thinking,鈥� she says.
Instead, the focus should be on what students are expected to learn and in which contexts it matters whether AI is used or not.
The seminar also addressed the energy use and environmental impact of generative AI. Knowledge in this area is still limited, but according to the speakers it is important to include this perspective in the discussion. In Lund University鈥檚 policy, sustainability is therefore highlighted as a principle, meaning that both energy use and broader sustainability aspects should be considered when deciding how the technology is used. The topic will also be explored further in the seminar series. On Thursday 19 November, Katherine Harrison will host a seminar on sustainability questions in relation to generative AI use, featuring Dr Ana Valdivia from the Oxford Internet Institute.
Both LiU and Lund University emphasise that work on generative AI is an ongoing process. Policies and guidelines are only a starting point, and it is through conversations and everyday practice that approaches take shape. The seminar series will continue throughout the year, providing further opportunities for discussion.
鈥淭his is not about finding a single correct answer, but about making informed and responsible choices in a rapidly changing reality,鈥� says Karin Axelsson.
The seminar series webpage provides details of the autumn seminars, as well as opportunities to join the mailing list and suggest topics for future events.
