Sociotechnical Considerations for Accessibility and Equity in AI for Healthcare

Abstract

As AI systems are built and deployed to support mental health services, it is imperative to fully understand the stakeholder acceptability of such systems so that these concerns can be taken into account in system design. As such, we undertook a consultation with staff (therapists) and service-users at Adferiad Recovery (a large mental health charity). The aim was to capture insights about their understanding of trust, and different trust factors for AI in mental health care. Surveys, interviews and focus groups were conducted with service users and therapists. Key takeaways for computer scientists and the developers of AI systems are presented.

Ähnliche Arbeiten