AI Ethics for Disability Justice

Summary

The study of disability and artificial intelligence (AI) from the viewpoint of disability studies and disability activism offers an important corrective to scholarly and historical trends that have ignored the unique needs, desires, and knowledges of disabled people. Various systemic issues and standards, from ableism to academic research barriers, have prevented disabled people’s perspectives from being fully considered in AI research. This absence is concerning because of the ways that AI, and digital technologies more broadly, may seriously impact, for better or worse, the quality of life for disabled people. Compounding this urgency, in the early 2020s, AI gained new widespread and global traction as large language models, such as OpenAI’s ChatGPT and Google’s Gemini, reached novel heights of advancement and application. These developments sparked a set of wider debates in scholarly, policy, and activist spheres about how to regulate, design, and use AI in ways that do not reproduce uneven power dynamics and inequality. Before the wave of AI hype in the 2020s, more critical scholars argued for more attention to how AI systems are bound up in material and ideological processes of oppression, violence, and extraction. As such, how AI is imagined, designed, regulated, and used has significant implications for all people and their proximity to justice. But, compared to the interrelated experiences and categories of race, gender, and sexuality, the experiences and knowledges of disabled people in AI ethics debates remain less represented.

Multiple fields of study, such as critical design studies, human-computer interaction (HCI) studies, AI ethics, critical AI studies, information studies, and new media studies, have sought to address this scholarship gap. Applying the framework of Disability Justice to AI systems and debates in particular helps demonstrate the importance of addressing the systemic ableism at the core of many processes, effects, and concepts at work in the full life cycle of AI systems. Disability Justice was developed by disabled activists of color and/or LGBTQIA+ orientations in North America. Disability Justice understands disability and ableism as inherently compounded by and in relation to other systems of oppression, including race, class, gender, sexuality, body weight, and nationality, to name a few. It emphasizes transformative justice and, therefore, can help identify and address the root causes of harm done by building and maintaining AI systems.