An empirical ethics study of the introduction of AI techniques to the assessment of cognitive decline (360G-Wellcome-213579_Z_18_Z)

The proposed research will examine how ethical concerns arise in the development of AI tools for screening and diagnosis and how and by whom they are addressed, focusing on the example of dementia. Further, it will examine the range of personal data on which such applications draw, and consider whether and how this affects what constitutes health data. The study proposes that applications of AI to dementia screening and diagnosis represent a particularly important case for investigation, given considerations about the nature of informed consent, and sensitivities associated with diagnosis, its communication and implications for individuals and families. The study will involve qualitative and ethnographic research with experts and members of the public to Identify applications of AI to dementia research, and outline central social and ethical concerns; Examine how ethical considerations are being negotiated in the development of AI tools ; Explore expert, public and patient perceptions of what constitutes reasonable use of different forms of personal data; Conduct a workshop to expand discussion and support future project development The project will result in outputs for clinical, policy and social science audiences, and form the basis for applications for funding to develop this work further.