Disability Data Justice from the Ground Up

Abstract

Disability advocacy organizations need access to disability data to advocate for human rights, disability justice, and sustainable development goals. Health informatics and Artificial Intelligence that has been developed through an equity-focused process provide important tools that can help address these needs. AI is prone to bias and might exacerbate discrimination and data ableism. A critical disability lens and community collaboration can help to address these biases, and multidisciplinary collaborations with grassroots and community-based organizations are crucial for advancing disability data justice. We have been engaged in a practice-led approach to building a disability justice-focused AI search engine. In the first section of this paper, we report on our co-design process, which included community consultations about disability data justice with local, provincial, national, transnational, and supernational disability organizations and advocates. In the second section, we demonstrate how we applied a transnational disability studies framework during our process of training a search engine AI to function from a disability justice perspective. We demonstrate the semantic and conceptual differences between a transnational approach and a disability rights approach, as a concrete example of how AI bias emerges. We argue that our participatory approach allows us to experiment with data repositories and search engines that confront AI bias and data ableism and reflect community needs.