A Virtual Collaboratory on AI and Justice led by York University will bring a new approach to equity-driven AI research.
The multi-institutional research hub will explore how ethical, political and social conditions shape AI design and governance – and how those foundations can be transformed to advance justice.
The collaboratory builds on a three-year trajectory of work at York by Faculty of Health professors Christo El Morr and Rachel Gorman, including a project on the use of AI for disability advocacy and the creation of open resources that support inclusive AI scholarship.
Those projects, funded by the Social Sciences and Humanities Research Council of Canada, culminated in the Equity & Artificial Intelligence International Symposium held at York, Nov. 6 and 7. The event, organized alongside co-applicants Elham Dolatabadi (Faculty of Health) and Laleh Seyyed-Kalantari (Lassonde School of Engineering), brought together more than 30 presenters from 20 institutions across five countries to examine the social, political and ethical stakes of AI.
AI in health care isn’t neutral, says Dolatabadi, and results are shaped by existing inequalities, including race and gender, because data reflects historical and social power dynamics.
“We must design with patients, not just for them, and give them space to advocate for themselves so that fairness principles are grounded in lived realities,” she says.
If a model doesn’t work well for certain patient groups, it points to issues in the data that go beyond technology, notes Seyyed-Kalantari. To address this, she says models need to be carefully tested to understand where bias exists and systems need to be built to stop those biases from continuing.
El Morr, symposium lead and facilitator of the new collaboratory, says, “We work toward AI futures grounded in liberatory foundations. Justice is not an afterthought. This work is about building intellectual, ethical and political infrastructures that make liberation thinkable in the design and governance of AI.”
Panels at the symposium addressed topics ranging from health equity and techno-ableism to governance and trust. Across sessions, speakers highlighted the limits of conventional fairness metrics and the need to transform the social conditions that influence data and design.
Insights from the international, interdisciplinary knowledge-sharing symposium helped to inform and define the mandate of the new collaboratory.
Serving as a digital and conceptual meeting space for scholars and practitioners, the new collaborative hub will advance equity-driven AI research by hosting working groups on governance models, data ethics and community partnerships. The goal is to amplify perspectives historically excluded from AI development, rather than approaching AI equity as a technical fix.
“The question is not how to fix bias inside a model,” says Gorman. “The question is whether these systems are already built on extractive relations. Justice-oriented AI requires reimagining the social conditions of knowledge-making.”
By linking initiatives across institutions and disciplines, the collaboratory aims to drive research on AI for justice, accessibility and social transformation by focusing on these priorities as foundational.
The Virtual Collaboratory on AI and Justice will launch in January 2026.
With files from Christo El Morr