Evidence from the frontline: Mental health in crisis affected contexts, episode 6: Artificial Intelligence and mental health chatbots
Listen to the podcast here or subscribe to Evidence from the frontline: Mental health in crisis affected context on Apple Podcast, Spotify, or wherever you find your podcasts.
In this sixth and final episode of Evidence from the frontline: Mental health in crisis affected contexts, we’ll be talking about Artificial Intelligence and mental health chatbots. Is it “too far, too fast?”
An AI chatbot is a computer program that uses artificial intelligence to simulate human conversation. While some research suggests that AI chatbots may show promise as digital mental health interventions, there is very little evidence on the effectiveness of chatbots developed using AI, especially more recent generative AI.
In this sixth episode, Sarah Harrison, Director of the Red Cross Red Crescent Movement MHPSS Hub, and Dan Amias, Senior Innovation Learning Adviser at Elrha, speaks with Mercy Githara, MHPSS Manager at the Kenya Red Cross Society, and Anne de Graaf, Technical Officer at the World Health Organization (WHO). They learn more about two chatbots: STARS, a non-AI chatbot developed by WHO, and Chat Care, an AI chatbot being deployed in Kenya by the Red Cross.
While chatbots may show promise providing services for young people and those with less serious conditions, complementing wider programming as in Kenya, we learn that the use of AI chatbots in humanitarian contexts carries significant risks. Cases like the ‘Karim’ chatbot, and lawsuits reported in the global media concerning generative AI chatbots and people at risk of suicide, underscore that humanitarian actors must strengthen our knowledge and capacity before we go too far, too fast. AI can strengthen interventions safely, such as through staff training and supervision, without coming into direct contact with vulnerable users. Sarah and Dan emphasise the need to build a shared evidence base, develop standards, crisis protocols and frameworks, seek guidance from practitioners and partners, and assess risks and costs (including environmental costs), before wider implementation.
Related resources
Key resources for practitioners
- Kenya Red Cross Society: Chat Care, AI Powered Mental Health Chatbot
- World Health Organization (WHO): Doing What Matters in Times of Stress evidence based self-help guide
- World Health Organization (WHO): Scalable Technology for Adolescents and youth to Reduce Stress (STARS) chatbot for more information contact: [email protected]
- Red Cross Digital MHPSS pledge: Leveraging and facilitating technology-empowered pathways (April 2025)
MHPSS Hub resources on Suicide prevention:
Find infographics, videos, podcasts, guides and tools on suicide prevention in the MHPSS Hub resource library: suicide - MHPSS Hub
Read more about the research:
- STARS chatbot(non-AI): de Graaff A.M, et al. (2025). Evaluation of a Guided Chatbot Intervention for Young People in Jordan: Feasibility Randomized Controlled Trial JMIR Ment Health 2025;12:e63515 63515
- Keyan, D., et al. (2025). The development of a World Health Organization transdiagnostic chatbot intervention for distressed adolescents and young adults. Front. Digit. Health 7:1528580. doi:10.3389/fdgth.2025.1528580
- Akhtar, A., et al. (2024). Scalable Technology for Adolescents and Youth to Reduce Stress in the Treatment of Common Mental Disorders in Jordan: Protocol for a Randomized Controlled Trial. JMIR Res Protoc 2024;13:e54585. doi: 10.2196/54585
- Karim Chatbot: Madianou, M.(2021). Nonhuman humanitarianism: when “AI for good” can be harmful. Information, Communication& Society, 24(6),850–868
- Torous, J. et al.(2025) Assessing generative artificial intelligence for mental health. The Lancet, Volume 406, Issue 10504, 683
- Spencer SW, Masboungi C. (2025). Enabling accessor automating empathy? Using chatbots to support GBV survivors in conflicts and humanitarian emergencies. International Review of the Red Cross.
Cooperation on developing digital MHPSS resources:
Please reach out by mail to the MHPSS Hub: [email protected] or Elrha: [email protected] if you, your National Society, organisation, university, or other stakeholders are interested in partnering or cooperating on researching into, and developing digital MHPSS approaches further.
Evidence from the frontline: Mental health in crisis affected contexts is a six-episode podcast mini-series produced in a collaboration between the MHPSS Hub and Elrha.
Designed for MHPSS practitioners working in humanitarian and crisis contexts, the series highlights impactful interventions and offers practical insights from experts in the field. It will explore critical topics including effectiveness, implementation, adaptation, integration into existing programmes, human resources, costs, and the importance of cultural and contextual relevance.