Adolescents Use GenAI To Co-Design DMHI Content
Researchers conducted two participatory workshops and focus groups in 2025 with 23 adolescents (aged 15–19) in rural Mpumalanga, South Africa, guiding use of ChatGPT, MidJourney, and Soundful to co-design stories, images, and music for the Kuamsha behavioral-activation app. Most participants found the process engaging (20/23), preferred AI-generated images (23/23) and music (19/23), reported mood improvements, but required prompt support and identified cultural biases, indicating promise tempered by bias and facilitation needs.
Key Points
- 1Show high engagement: 20/23 adolescents enjoyed GenAI co-design of stories, images, and music
- 2Reveal content preference: all participants preferred AI images; majority preferred AI music, indicating aesthetic resonance
- 3Indicate implementation needs: most required prompt support and flagged cultural biases, requiring facilitation and bias mitigation
Scoring Rationale
Demonstrates practical GenAI co-design benefits in LMIC adolescents, limited by small sample and persistent algorithmic bias.
Sources
Public references used for this report.
Practice with real Health & Insurance data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Health & Insurance problems
