Navigating the Future of Research: Insights from the REF-AI Report on Generative AI in Higher Education

Exploring the Impact of Generative AI on Research Practices

Addressing the Digital Divide in University Research

The Necessity of National Action for Responsible GenAI Use

Fostering Innovation Through Clear Guidelines and Integrity in Research

Navigating the Future of Research Assessment: Insights from the REF-AI Report

This blog was kindly authored by Liam Earney, Managing Director, HE and Research, Jisc.

The REF-AI report, a collaborative effort co-authored by Jisc and the Centre for Higher Education Transformations (CHET) and funded by Research England, shines a much-needed light on the transformative role of generative AI (GenAI) in research practices at universities. As we gear up for the next Research Excellence Framework (REF), the findings reveal that GenAI is already influencing approaches within the sector—some cautiously and some with inspiring innovation. Yet, much of this activity is happening under the radar.

Understanding the Current Landscape

For Jisc, the report’s findings resonate with our day-to-day experiences. We witness the unevenness of digital capabilities across institutions, where new tools often outpace the development of governance structures. The gap between emerging practices and established policy is significant, and it is a gap that the sector must collaboratively address. While UK Research and Innovation (UKRI) has rolled out guidance concerning GenAI in funding applications—focusing on honesty, rigor, transparency, and confidentiality—the guidance within the REF context remains vague. This ambiguity leaves institutions to navigate and interpret best practices alone.

Central to the report’s recommendations is integrity. The first recommendation stresses the necessity for every university to clearly publish its policy regarding the use of GenAI in research and specifically in REF-related activities. Establishing a transparent policy is essential to foster trust and build a fair assessment framework.

Unpacking Inconsistencies

The REF-AI report provides a window into real-world practices across UK universities. Some institutions are harnessing GenAI to sift through impact evidence, refine narratives, and even review outputs. Others are experimenting with bespoke tools tailored to streamline internal processes. While these initiatives stem from good intentions—aimed at managing rising workloads and the complexities of each REF cycle—they also lead to a patchwork of practices lacking clarity and shared understanding of GenAI’s role.

Many universities still lack formal guidance, and policy discussions are just beginning, leaving colleagues often unaware of how much GenAI has already woven itself into their institution’s REF preparations.

Bridging the Digital Divide

The findings also highlight a digital divide across the sector. Many academics express skepticism about GenAI’s role in REF, with some areas revealing disapproval rates as high as 70%. Yet, senior leaders show a burgeoning awareness that GenAI cannot be ignored. They recognize the responsibility to figure out how to utilize this technology effectively.

At Jisc, we understand that GenAI literacy varies, along with overall digital capability. Our mission is to help universities navigate these challenges. Just as we’ve implemented an AI literacy program for teaching staff, we must extend similar support to research staff, as highlighted by the REF AI findings.

The Importance of National Coordination

If we allow GenAI use to remain solely a local endeavor, we risk exacerbating the digital divide between institutions that can invest in customized tools and those that cannot. The effectiveness of a national research assessment exercise hinges on addressing these disparities.

Furthermore, establishing research integrity is foundational for the path ahead. Transparency must be at the forefront of any future initiatives. This is why the first recommendation—defining permissible GenAI uses and requiring full disclosure of its involvement in REF work—is crucial.

Fostering a Culture of Responsible Innovation

Introducing guidelines does not stifle innovation; rather, it creates the conditions for responsible experimentation. Institutions are eager for secure, trustworthy GenAI environments that uphold data protection, confidentiality, and research ethics, while providing clarity on balancing efficiency with academic oversight. We must learn from the past where digital advancement often outpaced standardized practices.

The REF AI report offers the sector the evidence necessary to transition from informal to structured approaches. As we approach the next REF amidst financial pressures and technological transformations, GenAI presents opportunities to alleviate administrative burdens and enhance consistency—provided it is embraced transparently with a commitment to integrity.

Conclusion: A Call to Collective Action

This moment calls for collaboration. Universities need policies, panels require guidance, and shared infrastructure must be established to level the playing field rather than deepen existing divides. Together, as a sector, we can build a future where GenAI not only aids in research but does so within a framework of trust and transparency that ensures fair assessment of UK research. Let’s take proactive steps today to shape an equitable tomorrow in research assessment.