AI is on its way to transform how mental health services are delivered by helping mental health professionals to better predict, deliver services, track conditions, and develop new treatments.
When building AI for mental health, user involvement has been mostly sought during data collection and labelling suggesting a gap in user-centric dialogue. This means that the actual impact of AI models in terms of effectiveness for mental health; or use and acceptance by laypeople, remain so far speculative raising questions related to reliability of AI based mental health model in real-world, their interpretation, trust, and fairness.
This project aims to understand how to build AI-based mental health models that support appropriate level of understanding, act responsible so that people can trust them and use them in everyday lives. Moreover, we are also interested in exploring how AI and digital tools may have exacerbated health inequalities and how to design data-driven systems addressing it.
Partners: HealthWorks, Ada Lovelace Institute
People: Muhammad Umair, Dawn Branley-Bell, Caroline Claisse, Abi Durrant, and Lynne Coventry