AI-Powered UX Research Framework for HIV Care in Marginalized Nigerian Communities
Generative AI augments UX research to protect vulnerable populations in digital health.
A new paper by Emmanuel Oluwatosin Oluokun and seven co-authors from Bournemouth University, Federal University of Technology Akure, and others presents a Generative AI-augmented UX Research methodology tailored for digital health in regulatory contexts. The work focuses on MSM (men who have sex with men) and transgender individuals living with HIV/AIDS in Nigeria, a population facing unique psychosocial and legal vulnerabilities. The researchers argue that existing UX research methods fail to account for stigma, privacy concerns, and cognitive load in these communities. Their approach integrates co-design workshops, thematic analysis, and requirements engineering into a four-stage UXR process. The first stage uses AI to generate hypotheses from existing data. The second stage plans foundational research with ethical guardrails. The third stage employs 'Building Blocks' to systematically extract insights while minimizing harm. The final stage constructs stakeholder-specific Point of View narratives that translate findings into actionable design guidance.
The core output is a set of ten UXR Play Cards, each containing actionable tasks, AI-augmented approaches, and ethical guardrails tailored for marginalized populations. For example, these cards might include strategies for anonymizing user feedback, using large language models to probe sensitive topics without triggering trauma, or visualizing user journeys that account for fear of exposure. The methodology is replicable and stigma-aware, advancing human-centered design for digital health. The paper was submitted to arXiv on 29 May 2026 under Computer Science > Human-Computer Interaction and Artificial Intelligence. It provides a critical framework for responsibly deploying generative AI in UX research while protecting vulnerable end users. The work has implications beyond Nigeria—any digital health intervention serving marginalized groups can adopt this privacy-centered, low-cognitive-load approach to ensure both ethical rigor and actionable insights.
- Methodology uses Generative AI for hypothesis generation, insight extraction, and narrative construction across four stages.
- Produces ten theory-informed UXR Play Cards with ethical guardrails for MSM and transgender HIV care in Nigeria.
- Framework is designed to be replicable, stigma-aware, and privacy-centred for marginalized communities in regulatory contexts.
Why It Matters
This framework enables safer, more effective digital health UX for vulnerable populations by integrating AI responsibly.