Slides and other materials for: Generative AI Across Research Traditions: Uses and Ethical Considerations

Resource is two files:Slides introducing the themes in the paperA set of resources derived from other research, summarising the key concerns, including (1) Key principles and challenges re: AI in research; (2) Governance domains prompts for understanding the ethics ecosystem and how it relates to AI research; (3) an 'issue card' for mapping tensions and issues in use of AI; (4) a reviewer resource to support peer review of AI research; (5) a glossary. Research uses of generative AI are receiving significant attention, with both optimism for potential benefits, and concerns regarding both appropriate use and the ways in which systems may have long-range and indirect impacts on scholarship. Concerns include the extractive nature of large language models and their reliance on invisible labour, bias in underlying data and thus in models and their outputs, techno-solutionism that may detract from tackling challenges, and significant concern regarding the claimed efficacy of tools for research purposes. Despite calls for novel governance mechanisms, and an emerging literature in this space, there is a gap in detailed application of established research ethics principles to specific classes of use of generative AI in research. This paper addresses that gap through a configurative narrative literature review. It discusses a range of high level concerns relating to use of generative AI in research, highlighting nine tensions. Across design-oriented, quantitative, qualitative, and evidence-synthesis based research orientations key uses of generative AI are noted alongside brief considerations for research ethics. The discussion aims to support researchers in navigating the ethics of generative AI use across research orientations.