The GAIN project (Generative AI literacy and kNowledge) ran for two months, and examined Generative AI literacy guides in universities in the UK and South Africa, specifically examining how equity, diversity and inclusion (EDI) are featured.
The new Framework Project Report identified that 'high-tier' universities are far more likely to actually produce guidelines for the use of Generative AI than 'mid-' and 'low-tier' universities, with the distribution gap being wider in South Africa than the UK.
Where the guidelines do exist, the report noted that complex ideas like algorithmic bias were expressed through a series of materialist metaphors, explaining them through concepts like a mirror, household waste, or genetic inheritance.
The report also found that where Equality, Diversity and Inclusion (EDI) is discussed in the guidelines, some institutions opt for a more strongly politicized approach than others. The team found that EDI language in South African guidelines had a marked political undertone, with more frequent mentions of access, equity and fairness, whilst British guidelines were more ethics-driven and institutionalised, with frequent reference to bias and inclusion.
The report was co-authored by Professor Pamela Abbott, Professor Laurence Brooks and Dr Sharon Wagg from the School of Information, Journalism and Communication at the University of Sheffield, with Dr Reggie Raju and Dr Andiswa Mfengu from the University of Cape Town. Dr Mfengu’s involvement came as part of a mobility grant arranged through the University of Sheffield’s International Partnerships scheme.
Prof Pamela Abbott said: “The report exposed quite a number of grey areas in our understanding of approaches to the critical use of generative AI and even our understanding of various kinds of literacies associated with the Information field. By comparing AI literacy guidelines between South Africa and the UK, we also revealed that the contexts where Generative AI is being appropriated shape to a great extent our understanding of ethical and responsible use of these tools”.
Prof Laurence Brooks, Professor of Information Systems, said: "This report represents the first step in looking at the GenAI guidance from an EDI perspective and highlights the gap between how this is addressed by more and less advantaged universities across two different countries and cultures. We want to continue this research and work towards understanding how the different contextual understandings of EDI in AI literacy can inform a global framework for AI literacy."
In total, the report identified 10 key findings, as well as 8 open research questions to guide further work in this area. The research team has been awarded additional funding from the School of Information, Journalism and Communication to further explore and integrate their findings into future Generative AI guidelines in Higher Education, as well offer practical guidance on their implementation via workshops later this year with information practitioners and academics.