麻豆淫院

Since the emergence of generative artificial intelligence (GenAI), universities have moved rapidly from debating whether AI should be used, to determining how it should be embedded across learning, teaching and assessment (McDonald et al., 2024). New and have emerged in response. Recent analyses of institutional AI policies have highlighted growing attention to governance, integrity and the responsible use of AI (Illingworth, 2026; Uttley, 2026). Likewise, emerging scholarship has advanced understanding of student engagement with GenAI, AI-supported learning and curriculum integration (O’Dea et al., 2026; Peiris, 2026). Such developments have significantly strengthened universities’ capacity to implement AI. However, they remain largely focused on implementation. They tell us how universities might adopt AI, but far less about how institutions should govern educational innovation itself.

As AI becomes embedded across higher education (HE), more fundamental questions begin to emerge: What educational purposes should GenAI serve? Whose educational values shape institutional decisions? Who benefits from these innovations, and whose voices remain underrepresented? Perhaps most importantly: How can universities ensure their approaches continue to evolve as evidence, technologies and educational priorities change?

Rethinking AI through responsible innovation

‘Responsibility should be embedded throughout the innovation process rather than in retrospect after new technologies have been adopted.’

While universities have made significant progress in developing policies and guidance for GenAI, HE may also benefit from a broader conceptual framework for thinking about educational innovation. At this juncture, the Responsible Innovation (RI) framework proposed by Stilgoe et al. (2013) offers a useful point of departure. Developed originally within the context of emerging science and technology, RI argues that responsibility should be embedded throughout the innovation process rather than in retrospect after new technologies have been adopted. To support this, Stilgoe and colleagues propose four interconnected dimensions: anticipation, reflexivity, inclusion and responsiveness. These interconnected dimensions encourage universities to think differently about AI adoption (figure 1). Rather than focusing solely on implementation, these dimensions invite institutions to anticipate educational consequences, critically examine their assumptions, include those affected by AI-enabled change and remain responsive as technologies and educational priorities evolve.

Figure 1: Responsible educational innovation: Reinterpreting Stilgoe et al.’s (2013) Responsible Innovation framework for institutional AI adoption in HE

Concluding thoughts: Towards responsible educational innovation

Universities have responded rapidly to the emergence of GenAI. Yet as AI continues to evolve, perhaps the most pressing questions facing HE are no longer technological – they are educational. The challenge is no longer simply how AI should be used, but why universities are embedding it, whose educational values shape these decisions, and what kinds of learning they seek to cultivate.

Perhaps the greatest risk is not that universities adopt AI too quickly but that they do so without sufficient institutional reflexivity. In the pursuit of innovation, efficiency and competitiveness, there is a danger that universities become preoccupied with what AI can do rather than asking what HE is for.

This is where Responsible Educational Innovation offers a useful way forward. By embedding anticipation, reflexivity, inclusion and responsiveness into institutional decision-making, it shifts attention from implementing AI to educational change. In doing so, it builds on the growing conversation within academia on the pedagogical use of GenAI, authentic assessment and academic integrity, while extending that discussion towards the institutional governance of educational innovation. Readers may also be interested in related 麻豆淫院 Blog posts on using GenAI pedagogically through graduate learning outcomes, generative AI in academia and authentic assessment and academic integrity in the age of GenAI. The question is no longer whether universities can innovate with AI, but whether they can innovate responsibly.

References

Illingworth, S. (2026). What UK university AI policies actually do: A study of 96 institutions. Higher Education Policy Institute.

McDonald, N., Johri, A., Ali, A., & Hingle, A. (2024). Generative artificial intelligence in higher education: Evidence from an analysis of institutional policies and guidelines (Version 1). arXiv.

O’Dea, X., Chiu, Y.-L. T., Bale, R., & Rossiter, M. (2026). Exploring student-AI interaction in assessment through the lens of learner agency: Case studies in UK higher education. Studies in Higher Education. Advance online publication.

Peiris, M. (2026). AI, affective development, and the ‘third space’ of learning. Journal of Learning Development in Higher Education, 40.

Stilgoe, J., Owen, R., & Macnaghten, P. (2013). Developing a framework for responsible innovation. Research Policy, 42(9), 1568–1580.

Uttley, L. (2026). Review and analysis of UK institutional guidance for artificial intelligence use in research. UK Committee on Research Integrity.

Artificial intelligence declaration: The author used Grammarly for spelling and grammar and ChatGPT for refining tone. The author takes full responsibility for the accuracy and originality of this publication and complied with the terms of .