Generative AI for Mental Health Support for Africans with Long-term Health Conditions

Mental health care in sub-Saharan Africa remains critically under-resourced, particularly for people living with long-term conditions (LTCs). Many countries operate fewer than ten public mental health institutions for populations of several million, a structural deficit compounded by socio-cultural stigma and limited mental health literacy that delays help-seeking and reinforces negative attitudes. These challenges often extend into African diaspora communities in countries such as the United Kingdom, contributing to crisis-driven service use and broader public health pressures.

To address these gaps, this study examines the potential of generative artificial intelligence as an accessible and scalable approach to culturally responsive mental health support. A mixed-methods study will be conducted with people living with long-term conditions and mental health professionals, incorporating ecological momentary assessments, semi-structured interviews, and distributed co-design workshops. These activities will inform the design of a culturally sensitive chatbot prototype and a custom GPT designed to deliver personalised support. The custom GPT will be deployed via an API on widely used instant messaging platforms and evaluated through a mixed-methods feasibility study, assessing usability, cultural fit, engagement, and perceived clinical value.

This project contributes to digital health and applied artificial intelligence by generating empirical evidence on the responsible adaptation of generative AI to culturally specific mental health contexts. The findings will inform a practical lightweight toolkit to support the ethical and culturally sensitive deployment of generative AI for mental health care among Africans living with comorbidities.

Research Questions and Objectives

This project, situated at the intersection of Human-Computer Interaction and Artificial Intelligence, seeks to generate insights into the digital and mental health needs of African populations with LTCs. The project will investigate these core questions:

1. How do socio-cultural factors impact mental health care?

2. What are the digital and mental health needs of Africans with LTCs?

3. How can generative AI be designed to support mental health in culturally sensitive and relevant ways?

4. Can culturally tailored mental health technologies effectively support Africans with LTCs in SSA and the diaspora?

This project is organized around three interconnected stages which will:

1. Investigate the socio-cultural determinants affecting mental health among people with LTCs, identifying barriers and facilitators to technology adoption in mental health support.

2. Develop scalable, culturally sensitive digital mental health solutions.

3. Evaluate these digital solutions to understand how socio-cultural factors influence mental health outcomes and create a toolkit for the ethical and culturally appropriate use of generative AI in mental health support.

Expected Contributions

This project is expected to yield both empirical and artefact contributions. Empirical contributions include insights into shared and unique mental health needs of African and Afro-diasporic populations in the UK, as well as knowledge of user requirements, workflows, and cultural considerations for generative AI development and use. Artefact contributions include the chatbot prototype, the fine-tuned custom GPT, and a toolkit for the responsible use of generative AI in mental health for African populations.

This project is sponsored by the Engineering and Physical Sciences Research Council (Grant number: EP/W52458X/1)