This month we had our first community meet up for the AI in Research Community. Members came together for an engaging and thought-provoking meetup focused on how AI is reshaping research practice. This month’s discussion explored both the ethical implications and practical challenges of using AI tools in scholarly work. From questions of peer-review integrity to the legal and labour issues shaping AI adoption, the session reflected a shared determination: to build a research environment where AI can be used responsibly, transparently, and equitably
Overview
The conversation opened with reflections on how AI is already appearing in key areas of the research lifecycle, from writing support to data analysis and peer review. Participants noted that while these tools hold enormous potential, their integration also raises difficult questions about authorship, bias, and accountability.
Key Discussions:
AI in Peer Review Raises Ethical Concerns.
Participants discussed growing worries about reviewers using AI to assess submissions. While AI may help with early drafting or summarisation, there was consensus that human oversight remains essential to preserve academic integrity. It was also noted that the use of AI in peer review could create inconsistencies in quality and tone, particularly if reviewers rely too heavily on automated text or interpretation tools. The group agreed that clearer boundaries and transparency about when AI is used during review would benefit both reviewers and authors.
Legal Risks in AI Usage.
The legal implications of uploading research materials into AI tools were explored, highlighting the need for clear institutional guidance on rights clearance and data protection. Participants discussed the risk of inadvertently breaching copyright or confidentiality agreements when sharing unpublished or sensitive data with AI systems. This led to calls for sector-wide frameworks to clarify acceptable practices and responsibilities when engaging with third-party AI platforms.
AI’s Role in Pre-Peer Review Processes.
It was suggested that AI could enhance the quality of submissions before formal review, helping reduce reviewer workloads and improving efficiency in publication workflows. For example, AI could support researchers by checking structure, clarity, and adherence to journal requirements, or even flagging potential ethical or methodological gaps before submission. However, participants cautioned that such tools must be transparent and accountable to avoid introducing bias or false confidence in weak papers.
Pressure on Academics Driving AI Adoption.
The group reflected on how workload pressures and time constraints often push researchers to adopt AI tools. This prompted a broader discussion about the systemic factors driving such reliance and the need for more sustainable working models. Participants acknowledged that while AI can save time, it may also mask deeper issues of overwork, unrealistic expectations, and inequity in academic publishing. Addressing these pressures at an institutional level was seen as key to ensuring ethical and balanced AI use.
Opacity of AI Tools Undermines Trust.
Participants emphasised that without transparency about how AI systems generate outputs, reproducibility and trust in research outcomes could be compromised. The “black box” nature of many AI tools makes it difficult for researchers to understand or explain how results are produced, which poses challenges for verification and accountability. Greater transparency and documentation from tool developers were identified as essential steps toward responsible AI adoption in research workflows.
Need for Sector-Wide Consensus on AI Use.
There was strong agreement on the need for shared principles and guidelines that define responsible AI use in research across the sector. Participants stressed that isolated institutional policies create confusion and inconsistency, whereas sector-level coordination could build confidence and alignment. The goal is to establish norms that protect research integrity while allowing innovation to flourish responsibly.
Doctoral Researchers Need Clear Guidance.
Attendees noted that doctoral researchers often lack tailored advice on using AI tools and called for policies that meet their unique needs while aligning with broader institutional approaches. Early-career researchers are frequently experimenting with generative AI for literature reviews, writing, or coding, but many remain unsure about what counts as acceptable use. Participants agreed that specific training and mentorship in this area would be invaluable.
Training Supervisors on AI Policies.
The role of supervisors in modelling ethical AI use was discussed, with participants emphasising the importance of clear training and consistent institutional messaging. Supervisors play a key role in shaping research culture, so ensuring they understand both the risks and benefits of AI tools is vital for cascading best practice throughout academic departments.
AI’s Impact on Research Reproducibility.
The potential impact of AI’s opaque processes on reproducibility was identified as a key concern, underlining the importance of transparency and documentation in AI-assisted research. Participants discussed how hidden data sources or model biases could undermine confidence in findings, particularly in fields where reproducibility is already challenging.
Transcription Tools and Data Privacy Risks.
The group discussed risks associated with AI transcription tools, including issues around data storage and privacy, and stressed the importance of evaluating tools before use. Researchers were encouraged to check where their data is processed, whether it is stored externally, and if any information could inadvertently breach confidentiality agreements.
Frameworks for Evaluating AI Tools.
Several participants shared examples of frameworks being developed within institutions to assess AI tools for safety, licensing, and appropriateness for research use. Such frameworks help ensure that decisions about adopting AI are grounded in ethical, legal, and operational awareness rather than convenience alone.
Continuous Communication is Key.
Policies alone were seen as insufficient; participants agreed that ongoing communication, training, and dialogue are vital for embedding responsible AI practices across the sector. Regular updates, case studies, and shared experiences were seen as practical ways to ensure that guidelines evolve alongside technology and remain relevant.
Jisc’s Role in Supporting the Community.
The session closed with a reminder of the importance of sharing community needs and feedback to help shape Jisc’s continuing support for the AI in research ecosystem.
Closing Thoughts
The meet up underlined a shared message: AI in research isn’t just a technical challenge…it’s a cultural one. As universities and research organisations adapt, the community’s collective efforts to establish transparent, fair, and inclusive frameworks will determine how successfully AI supports rather than disrupts scholarly and academic integrity.
We look forward to continuing the conversation at our next community event, which will be taking place next week on Monday the 3rd November and welcome all to join the conversation in shaping what responsible AI in research truly looks like.
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