On 18 February, more than 30 academics and postgraduate researchers from across BBS and BLS came together for an engaging, hands-on workshop exploring how Generative AI (GenAI) can support different stages of the research process. Hosted by Professor Noelle Quénivet and Dr Sarah-Louise Weller, the session was designed to be practical and reflective. It was suitable for colleagues at all levels of experience with GenAI, whether just beginning to experiment or already integrating these tools into everyday research practice.
The workshop set out to demystify GenAI, highlight responsible and ethical use and offer real examples of how colleagues across the College are already using AI tools to streamline their research workflows.
Understanding the landscape: what GenAI can and can’t do
The session opened with an informative presentation from Boudewijn Dominicus (Research Development Manager, Health & Science) and Lindsay McDonagh (Research Development Manager, Creative Technology). They introduced the capabilities of GenAI tools and UWE Bristol’s newly issued Guidance on Generative AI and Large Language Models in Research.
Their talk covered how large language models (LLMs) work, accuracy challenges and common pitfalls, issues around data security, ethics and environmental impact and how funders are responding. They also outlined the policies researchers need to be aware of. UWE’s policy for using AI, specifically Microsoft Copilot, was introduced alongside a clear and memorable framework for effective prompting (RISEN):
R – Role: Give Copilot a specific role
I – Instructions: Provide clear and specific instructions
S – Steps: Break complex requests into manageable steps
E – Examples: Provide examples of desired outputs when helpful
N – Notes: Add important constraints and context
They also shared several practical “pro tips”, including asking the model to rewrite your prompt for clarity, refining prompts iteratively (“multi-shot prompting”) and specifying desired output format, tone or length. They highlighted ways Copilot can support the research funding process, from aligning ideas with call scopes to structuring narratives, editing, tailoring content for different audiences and preparing for interviews.
Exploring ethical questions through scenario-based discussion
Participants then worked in groups to unpack the ethical and legal dimensions of three fictional (but very realistic) scenarios involving GenAI use in research. These discussions encouraged colleagues to reflect on confidentiality, attribution, fairness, transparency and the responsible handling of data when using AI tools.
How researchers are using GenAI right now
One of the highlights of the session was a series of demonstrations from colleagues and PGRs across CBL. Each showcased a different real-world application of GenAI in research practice.
M. Emre Hayyar: Gemini as a learned friend
Emre highlighted the importance of clearly setting boundaries in prompts. Using a powerful “kill-switch” example, he showed how the quality and accuracy of outputs can differ dramatically depending on how a prompt is constructed. His take-home message was clear: LLMs tend to be overly supportive, so ask them to be critical or to take on the role of a supervisor, reviewer or journal editor.
Dr Mike Pollard: using LLMs to build research scenarios
Mike demonstrated how an LLM can act as a research assistant when given detailed and highly specific instructions. He emphasised the importance of fact-checking and refining outputs. He also showed how to instruct the model to exclude certain sources, such as particular years, languages or jurisdictions. He highlighted the need for researchers to reshape and refine generated material themselves to ensure it meets the needs of the task.
Dr Neeti Shikha: from manuscript to conference-ready slides
Neeti shared how she uses Gamma to transform a 5,000-word manuscript into a 15-minute conference presentation. The tool helps create visually engaging, audience-appropriate slides that can be adapted for academics, students or practitioners. She also offered helpful tips for responsible use.
Alexander Grimmig: from annotation to insight
Alex demonstrated how he integrates Zotero, Obsidian and AI tools to annotate, connect and synthesise his notes on journal articles and reports. This combined workflow helps avoid fragmented notes and reduces time spent re-reading papers.
To round off the afternoon, Professor Noelle Quénivet demonstrated how she uses Copilot to draft abstracts and conclusions for journal articles, as well as composite conclusion chapters for books. She closed with an important reminder to always check your target journal’s policy on AI use and disclosure before submitting.
A collaborative and insightful session
Huge thanks go to all the presenters for generously sharing their expertise and to all attendees for their thoughtful contributions and insightful questions throughout. The workshop showed how creatively colleagues are beginning to use GenAI, while also reminding us of the care, ethics and critical thinking required to use these tools responsibly.
The strong engagement across the Business School and Law School demonstrates a genuine appetite for learning how GenAI can enhance, streamline and support high-quality research. We look forward to building on this momentum with further sessions in the months ahead.
PS: For full disclosure, Copilot was used to structure notes taken during this three-hour session and the final version was edited using Copilot.
