Information Research

Vol. 30 No. 2 2025

Generative AI in higher education. The good, the bad, and the ugly

DOI: https://doi.org/10.47989/ir30251409

Pulk, Kätlin and Koris, Riina (Eds.). (2025). Generative AI in higher education. The good, the bad, and the ugly. Edward Elgar. 2025. 214 p. ISBN: 9781035326013

This timely collection features papers from a geographically wide range of countries. The editors are from Estonia, with contributions from the UK, Ireland, the Netherlands, Turkey, Denmark, Finland, China, the USA, and Australia. This welcome change from collections overly reliant on the English-speaking world highlights an important point: these systems predominantly use English language training materials, resulting in inadequate and biased responses to questions about history and culture in countries with less widely spoken languages (p. 9).

What unites these countries is generative AI’s immediate impact on teaching, learning, and research in higher education. All of them face similar challenges: advising on ethical use, assessing student progress amidst potential generative AI use, and preparing students for workplaces that will inevitably employ this technology.

The editors' introduction serves both as a chapter review and a valuable analysis of generative AI's positive and negative aspects. They highlight academic pressures as justification for technological assistance while noting potential impacts on human analytical, reasoning, and interpretative capabilities.

The introduction sets the scene, as do the next two chapters. In Chapter 2, Martin and Williams explore Dreyfus's potential response to generative AI, concluding that he would have emphasised its limitations and ‘taken mischievous delight in the tasks on which it spectacularly fails’ (p. 25). Chapter 3 presents a discourse analysis of academic interviews, revealing metaphors like ‘elephant in the room’ and ‘digital tutor’ that reflect concerns about balancing generative AI's advantages and risks.

The ‘Good’ section presents positive approaches to generative AI: Pavlik considers pedagogical benefits through constructivist learning, concluding that generative AI has potential that requires considerable effort to realise. Hendriksen examines generative AI as a personal tutor through Bloom's two-sigma problem lens (the finding that one-to-one tutored students perform two standard deviations better than those in conventional classrooms), suggesting these systems' diverse knowledge access may benefit education's growing focus on multidisciplinary challenges.

Chapter 6 explores generative AI as an enabler for educators, demonstrating applications for adding local context, enhancing course design, and aiding assessment. Kerem cautions that ‘effective utilisation depends on a thorough understanding of its capabilities and limitations’ (p. 86).

Finally, Dowling and Li discuss generative AI in academic research, showing systems can be fine-tuned for specific tasks and act as collaborators. For further collaboration, however, ‘many of the critical benefits will rely on higher-level GPT models than currently exist’ (p. 99).

Part III, the ‘Bad’ has four chapters and deals with ranges from Clark and Denman’s consideration of generative AI as a disrupter of creativity through its impact on assessment and the ethical challenges of bias and data privacy to a further analysis of the ethical pitfalls involved in using generative AI.

Finally, Part IV, the ‘Ugly,’ consists of two chapters: the first deals in greater depth than the previous chapters with the moral and ethical issues in the use of generative AI in higher education, and the last states that the advice given to academic staff on what they need to do in relation to generative AI (which is, generally, too many things without much guidance) may not be entirely ‘ugly,’ but certainly is not ‘good.’ The author, Michelle Miller, offers guidance on how this situation might be changed from the ‘ugly’ or ‘bad’ to the ‘good.’

The quality of the contributions in this volume is high: the authors review existing research and deal with both positive and negative aspects of generative AI use in higher education. From their experience, the contributors are able to offer original insights into the problems and practise of generative AI use, stimulating further thoughts on these issues by the reader. The reference lists at the end of each chapter are extensive and will serve to guide readers to related research and diverse experiences.

This book is an excellent introduction to the application of generative AI in higher education and should be available in all universities wrestling with these problems. The authors’ very diverse geographical backgrounds suggest that their advice has wide applicability. The authors all speak from a deep understanding of generative AI and from experience in its use by themselves or by their students. The only drawback is the price, although an e-book is available at a more modest £25.00.

Prof. T.D. Wilson

Professor Emeritus, University of Borås

April 1, 2025