Abstract

Artificial Intelligence (AI) for healthcare holds immense promise but carries a profound risk of amplifying existing health inequities, particularly for underserved groups like Pacific peoples in New Zealand. Standard AI models can perpetuate and scale social and clinical biases that lead to poorer health outcomes. This paper argues that to build equitable AI, we must move beyond purely technical fixes and adopt a new methodology grounded in community partnership. We propose an Equity-Centred Co-Design Framework that directly targets the sources of bias. Using the development of an AI-powered management system for Pacific heart failure patients as a case study, we demonstrate how this framework is applied. By integrating Pacific worldviews, our approach ensures that the community's lived experience shapes the AI's development from its foundation. This paper offers a practical roadmap for researchers and developers to create AI systems that are trustworthy, culturally responsive, and grounded in social justice principles.

Download

Share

COinS