Sheffield Researchers Win Impact Award at UK's First National AI Research Conference for Responsible AI in Practice
Sheffield team won the UK's first AI impact prize for making responsible AI workable in major organisations. Policy updates at the British Library and Council, plus PAC cite.

University of Sheffield researchers win UK's first national AI impact prize
Date: 24 September 2025
Academics from the University of Sheffield's Centre for Machine Intelligence have received the impact award at the UK's first national conference for AI research. Their work focuses on making responsible AI practical inside major organisations, including the British Library and Sheffield City Council.
Credit: UK AI Research Symposium
Why this matters for PR and Communications
- Proof that responsible AI can be translated into clear, day-to-day practices that staff can follow.
- A model for cross-functional governance and stakeholder engagement you can adapt for policy, reputation, and internal adoption.
- Immediate talking points for executive narratives on AI risk, ethics, and public trust.
What the Sheffield team did
Led by Dr Denis Newman-Griffis and Dr Susan Oman, the project brought together experts from Information, Journalism and Communication, History, Philosophy and Digital Humanities, and the University Library. Partners included the British Library; Sheffield City Council; Eviden; and the Open Data Institute.
- Reviewed existing AI policies across large, complex organisations.
- Spoke with staff in varied roles to learn how AI shows up in real work.
- Converted broad principles into concrete practices suited to each organisation.
The research gives leaders a practical route from policy statements to workflows, training, and governance that teams can actually use.
Results you can reference
- Policy updates at the British Library and Sheffield City Council.
- Recommendations cited in the Public Accounts Committee's report on the Use of AI in Government.
- A follow-up workshop with Sheffield City Council to support other local authorities in South Yorkshire.
For context on the event, the UK AI Research Symposium (UKAIRS), organised by Responsible AI UK, brought together researchers from across disciplines and presented its first and only prize to the Sheffield team.
How the researchers frame responsible AI (in practice)
Dr Denis Newman-Griffis highlighted that responsible AI is vital, yet hard to put into action without working closely with teams who are testing it on the ground. He noted that people, processes, and place all affect what "good" looks like, and that collaboration with partners made the work tangible for day-to-day decisions.
Dr Susan Oman emphasised that responsible AI means different things in different workplaces. Rather than forcing universal definitions, the project invites reflection on care, inclusion, and shared responsibility-and how language can slip without deliberate practice.
Creative engagement: Constant Washing Machine
The team collaborated with Blast Theory on an interactive artwork that helps audiences engage with responsible AI. Called Constant Washing Machine, it uses soaps engraved with phrases-such as care, shared meaning, practice, and inclusion-to show how everyday choices influence responsible use of AI, and how language can feel slippery in real contexts.
Credit: Constant Washing Machine by Blast Theory
What PR and Comms teams can do next
- Set the standard: Publish a plain-language AI use policy with examples of acceptable and unacceptable use across your channels and internal workflows.
- Own the narrative: Build a clear risk-and-benefit message house for AI. Prepare Q&A for media, staff, and stakeholders.
- Operationalise it: Tie policy to process-approvals, vendor checks, data guidance, and human-in-the-loop reviews.
- Train your people: Run role-specific sessions for copy, social, media relations, and design. Focus on ethics, attribution, consent, and disclosure.
- Measure and iterate: Track incidents, time saved, and quality outcomes. Review quarterly and update policy as tools and use cases change.
- Engage creatively: Use interactive formats (like the artwork approach) to make principles memorable for staff and the public.
About the event and partners
The prize was awarded at the UK AI Research Symposium (UKAIRS), organised by Responsible AI UK. Project partners included the British Library; Sheffield City Council; Eviden; and the Open Data Institute-reflecting a wide coalition across culture, local government, industry, and data policy.
Useful links
Level up team skills
If you are building practical AI literacy across PR roles-from policy to content and approvals-browse role-based options at Complete AI Training.