Stakeholders’ perceptions of AI-assisted obstetric ultrasound in rural and urban Kenya: a qualitative study

Abstract Introduction : As artificial intelligence (AI) tools gain popularity in maternal healthcare, understanding community and health system perspectives is essential for inclusive and ethical implementation. This study explores stakeholder views on AI-supported ultrasound for pregnancy care in Kenya, that emphasized trust, equity, system readiness, and the importance of community engagement. Methods : A qualitative study was conducted in two Kenyan settings: Nairobi (urban) and Kilifi County (rural). Fifty-five semi-structured interviews and three focus group discussions were conducted with pregnant women, partners, healthcare providers, Sub-County Health Management Teams (SCHMTs), Community Health Promoters (CHPs), and community and religious leaders. Data were analyzed inductively using grounded theory principles. Results : The participants viewed AI with cautious optimism. While many welcomed its potential to enhance diagnostic accuracy and access, concerns emerged around data privacy, informed consent, quality of care, and over-automation. Reliability or trust was linked to transparency, human oversight, and the use of local data. Participants highlighted structural challenges, including weak referral systems and staffing gaps, raising questions about the health system’s readiness to integrate AI. Fears of equity and “pilotitis”, short-lived projects without sustainability, were also expressed. Conclusion: Findings from this study suggest that AI implementation in maternal health may need to go beyond technology development and deployment. Developers and relevant stakeholders may need to consider co-designing with communities, alignment with health system realities, and the integration of ethical principles such as equity, accountability, and transparency.