- University of Sheffield academics have been recognised for their research that is driving the responsible use and development of AI
- Research has helped inform responsible AI policies for major organisations across the country, such as the British Library and Sheffield City Council
- Sheffield researchers were honoured at the UK's first national AI conference to champion the strength and impact of UK AI research
A new award recognising the strength and impact of the UK's AI research, has been awarded to academics at the University of Sheffield.
The prize, awarded at the UK's first national conference for AI research, has been awarded to academics from the University's Centre for Machine Intelligence for their research which is driving the responsible use and development of AI at major organisations across the country, such as the British Library and Sheffield City Council.
Led by Dr Denis Newman-Griffis and Dr Susan Oman, the researchers have been working with the organisations to shape best practice for using AI in their sectors. The team took a cross-cutting approach, bringing together academics from the University's School of Information, Journalism and Communication, the School of History, Philosophy and Digital Humanities, and the University Library.
Partners in the project include: the British Library, the UK's national library and a world-leading research and heritage institution; Sheffield City Council, a local authority serving over half a million people and a leader in digital transformation in local government; Eviden, a multinational digital transformation and AI firm; and the Open Data Institute, a leading non-profit organisation shaping policy and practice on data and AI.
As part of the project, the Sheffield academics analysed current AI policies and the perspectives of staff working with AI in a variety of roles in the organisations to understand what 'responsible AI' means in day-to-day practice and what makes it achievable.
The research provided much-needed insight into the rapidly-shifting AI landscape by addressing significant gaps in knowledge for how complex organisations can translate broad statements of AI policy into actionable practices that reflect their unique needs.
Findings from the study have already informed AI policies in the British Library and Sheffield City Council. Recommendations from the project have also been cited in the Public Accounts Committee's recent report on the Use of AI in Government, demonstrating the rapid impact of the research in shaping national policy. A follow-up workshop with Sheffield City Council has brought recommendations to other local authorities in South Yorkshire, informing ongoing AI adoption.
Dr Denis Newman-Griffis, Senior Lecturer and Theme Lead for AI-Enabled Research at the University of Sheffield Centre for Machine Intelligence, said: "Responsible AI is an enormously important area right now, but it's also one that can be quite difficult to grasp in practice. By working with partners who are actively exploring what good practice with AI looks like in their own organisations and wider sectors, we've been able to learn much more about how different people, processes, and places affect responsible AI.
"Working closely where the rubber meets the road with AI is vital for bridging the gap between responsible AI policy and practice, and our partners have shaped the research as much as we've helped inform their practice. Receiving this award is a wonderful recognition of the amazing work our team and partners have done to make responsible AI a bit more tangible and day-to-day."
As part of the collaboration, the researchers also worked with the multi-award-winning artists, Blast Theory, who developed an interactive artwork to represent and engage people with responsible AI use.
Called Constant Washing Machine, the artwork conceptualises responsible AI through the metaphor of soaps engraved with key phrases, such as 'care', 'shared understanding', 'practice' and 'inclusion'. It captures the everyday, social nature of the decisions that affect responsible AI in practice and the slipperiness of its language.
