The AI in research community is continuing to grow into a supportive network of people, sharing knowledge, conversations and experiences on AI across the sector. Our latest meet up again brought interesting discussion. particularly in relation to the REF (Research Excellence Framework), the ethical considerations of AI, institutional policies, and the challenges of integrating AI tools responsibly in research environments.
We discussed the challenges of using AI to assess research quality, the copyright concerns around training data for large language models, and the importance of transparency in how AI is used. The community also highlighted the need to improve AI literacy across the research community and explored institutional licensing options to support fair access to AI tools.
Key takeaways
AI in REF Report Concerns: The AI in REF report highlighted recommendations for managing AI in the research assessment process but raised concerns about increased administrative burdens rather than reducing them. Institutions are grappling with auditing AI usage and ensuring transparency.
REF 2029 Announcements: The recent REF 2029 announcements lacked explicit guidance on AI integration, leaving institutions uncertain about future directions. While data and technology were mentioned, AI-specific policies were notably absent.
Responsible Use of AI in Research: There is an emerging need for external mandates or sector-wide policies to ensure responsible and transparent use of AI tools in research. Without these, institutions face challenges with inconsistent practices and inequities.
Ethical Challenges with AI Tools: Discussions emphasised the importance of responsible metrics and policies to prevent misuse of AI in assessing research quality. Concerns were raised about biases favouring English-language outputs and the ethical implications of using AI for peer review.
Copyright and AI Training Data: The use of copyrighted materials to train AI tools remains a contentious issue. Cases like the Anthropic settlement highlight the need for better protections for authors and clearer policies around Open Access content.
Transparency in AI Use: Best practices for transparency in research outputs include acknowledging and describing AI usage, even if reproducibility is challenging. This aligns with broader efforts to promote open research principles.
AI Literacy in Research: Increasing AI literacy across the research community is critical to ensure ethical data usage, informed decision-making, and appropriate tool selection. Specialised tools tailored to research needs are emerging as alternatives to general-purpose large language models.
Institutional Licencing for AI Tools: Institutions like Oxford and LSE are providing access to specific AI tools (e.g., Claude) to reduce digital inequities. Jisc is exploring licencing options to make similar tools accessible across institutions.
Critical Thinking and Appraisal Skills: The importance of fostering critical thinking skills among researchers and students was highlighted as essential for evaluating AI-generated outputs and avoiding reliance on ‘AI slop’.
Going forwards
Looking ahead, some of the members of the community called for clearer national guidance on how AI fits into research assessment for the future, especially as a lot of the community remains vague on the role of REF 2029. Institutions should focus on developing consistent policies for responsible AI use, improving transparency, and strengthening AI literacy to support ethical and informed research practices. It is also a hope that collective licensing models through Jisc can contribute to reducing digital inequities by widening access to trusted AI tools. Overall, there is a shared recognition that coordinated action is needed to ensure AI supports rather than complicates the future of research.
Our next community meet up will take place on Wednesday the 21st of January 3:30pm – 4:30pm. We welcome all to join the conversation in shaping AI and research. A timely reminder will be sent along with a link, in the new year in our upcoming newsletter.
Find out more by visiting our Artificial Intelligence page to explore publications and resources, learn more about our communities and sign up for our AI Literacy training.
For regular updates from the team sign up to our mailing list.
Get in touch with the team directly at AI@jisc.ac.uk