Beyond AI as Tool: Indigenous Knowledges and Relational Accountability in Educational Technology

Abstract

This paper emerges from the University of Victoria-hosted funded research project Meta-Relationality and AI: Discernment, Fields, and Relational Capacity in AI Systems. The project examines whether generative AI systems can be redirected away from reductionist reasoning through small shifts in ontological framing. I bring this inquiry into conversation with Indigenous knowledges, including First Nations teachings about interdependence, to ask how AI might be approached through relational accountability rather than instrumental use. Dominant AI adoption in education often frames the technology as an assistant, tool, or productivity system. These framings tend to assume separability: the idea that humans are separate from nature, and that intelligence, learning, labour, land, energy, water, and consequence can be treated apart from one another. Indigenous onto-epistemologies, understood here as situated ways of knowing and being, interrupt this assumption by foregrounding relation, responsibility, land, ancestry, reciprocity, and the more-than-human conditions of learning. I propose meta-relationality as a pedagogical orientation that can help educators and learners notice the costs, dependencies, and consequences that are often externalized in AI discourse and practice. The paper offers guiding questions for AI education grounded in relational accountability, with attention to the risks of extraction, appropriation, epistemic containment, and technological solutionism. Rather than asking how Indigenous knowledges can be added to AI ethics, I ask how Indigenous knowledges and meta-relationality can interrupt the deeper architectures of separability that shape educational technology itself.