Citizen science and negotiating values in the ethical design of AI-based technologies targeting vulnerable individuals

Abstract Citizen science is the new mantra both in academic circles and the public discourse. While the citizen science ideal is conceptually broad, If and how it can be realized in fields often depicted as value free/value neutral—such as applied AI—is controversial. The practical challenges in generating ethical AI encapsulating the citizen science ideal are addressed by targeting scientific practices underlying the participatory design of an AI-based tracking app aimed at enhancing the safety and wellbeing of vulnerable citizens with dementia of a Danish municipality through the engagement of the local community. The focus is on the process of social value construction and its citizen science rationale: how design values have been debated, traded-off, and selected via participatory-deliberative methods and by engaging experts and non-expert stakeholders as citizen scientists. An emphasis is on the import of dialogic interaction for negotiating values in open conversations within diverse groups of interest. Deliberative procedures are beneficial to produce ethical AI embodying vital citizen science desiderata since users’/citizens' values, needs, and expectations are fulfilled while technical-efficiency standards are also met. The result is a public methodology for designing ethical AI that better expresses the true spirit of citizen science in liberal democracies (value-laden, pluralistic, inter-disciplinary, inclusive, participatory, cooperative, solidarity-oriented). Hence, trust and acceptance of the AI generated, even of contentious “surveillance” technologies, is enhanced since the digital innovation can be perceived as truly citizens-/humans-centred and society-oriented.