AI Ed Wiki logoAI Ed Wiki
Use with AI

📄 Research Article

Reorienting Critical AI Literacy: A Community-Rooted Praxis of “Resisting AI”

● high · 2026-08-16 · critical-ai-literacy, critical-pedagogy, equity, ai-education, community, agency, reducing-ai-misuse, ethics

Synthesis: Li & Mroziak (2026) propose adding "Resisting AI" as a necessary stance within Critical AI Literacy (CAIL). Critiquing a dominant discourse that assumes AI inevitability and techsolutionism — and whose productivist, extractivist logic renders communities powerless — they argue CAIL should build collective agency through dialogic, collaborative, community-rooted pedagogies. Drawing on Freirean praxis, grassroots and transnational organizing, and the roles of educators and community organizers, they position education as a space where communities imagine and build alternative technological futures.

Key Findings

CAIL is needed but currently reactive, not contesting. Recent scholarship maps critical perspectives on AI across five emerging themes — impact, ethics, interventionism, values and power, and interdisciplinarity — with shared concerns over coded biases and disproportionate affordances. Yet the authors argue that the existing framing of CAIL "remains largely reactive rather than contesting," presuming that AI is and will be needed and that it is the sole solution to pressing global challenges. This positioning is central to the AI literacy debate about whether literacy frameworks reproduce rather than challenge dominant power relations.

The dominant discourse assumes inevitability and techsolutionism. Across more than 600 policy frameworks and guidelines — the EU's Digital Education Action Plan, UNESCO's AI Competency Frameworks for students and teachers, the OECD AILit framework — AI is routinely treated as inevitable ("AI is here to stay") and as the ultimate solution to political and environmental problems. The authors critique the "enchanted determinism" of international organizations, noting that determinist, market-driven undertones disorient the framing of Ethics and Agency within CAIL: agency becomes highly individualized, and ethics risk reducing to symbolic gestures that favor industry over community concerns.

Resisting AI reframes CAIL through Freirean praxis and conscientious refusal. Grounded in Paulo Freire's concept of praxis — the synchronization of reflection and action — the authors conceptualize Resisting AI as a pedagogical orientation rather than a fixed checklist. It harnesses two forms of refusal: practical refusal (asking "Question Zero" — is AI needed to begin with? — making opting out viable and contesting data inputs and outputs) and ideological refusal (diversion from the extractivist, productivist logic wired into commercial AI, which replicates colonial patterns of hegemonic extraction).

Resisting AI is a pluralist, community-rooted praxis. The paper enumerates grassroots, community-led initiatives in intersectional and transnational contexts: planetary-justice advocates co-designing art-based campaigns against data-center expansion; gender and LGBTQ+ communities organizing around encoded cisheteronormativity; Latin American data workers building informal networks of care and collective resistance around digital labor; and decolonial practices of unlearning, co-learning, and relearning. Together these exemplify participatory, situated processes that build collective agency and pluralize CAIL.

Educators and community organizers play distinct but complementary roles. Educators — especially in K-12 and universities — act as "mediators of power" who cultivate intentional pauses, design learning experiences that include non-use and refusal, and legitimize refusal and uncertainty as thought-provoking outcomes. Community organizers, grounded in lived experience and relational accountability, frame refusal not as withdrawal but as a collective act of world-building — surfacing experiences of surveillance and data harm and producing community-authored knowledge (statements, toolkits, zines).

Resisting AI insists education remain a space for building alternative futures. It is not an anti-technological stance but a commitment to collective self-determination: education should remain a space where communities decide, together and in relationship, what kinds of technological futures are worth inhabiting. Its essential components are the acumen to discern and critique AI discourse and usage; the intellectual habit of practical and ideological refusal; and pluralist epistemology, dialogic pedagogies, and radical imagination for community flourishing. This reframing positions resistance and non-use as legitimate educational outcomes rather than merely as misuse to be policed.

Study Design & Method

This is a perspective/conceptual paper (RESPECT 2026), not an empirical study. It proceeds through a critical review of existing AI literacy frameworks, guidelines, and policies (EU, UNESCO, OECD, AI4K12, and others), a positionality statement, and a synthesis of Freirean critical pedagogy, data justice initiatives, and documented grassroots organizing across intersectional and transnational contexts. It concludes with roles for educators and community organizers and a set of limitations and openings for future practice.

Implications for AI in Education

The paper directly challenges the instrumental, solutionist framing of AI literacy that dominates many curricula and policies. It connects AI education to critical pedagogy and collective Agency, arguing that literacy instruction should equip learners not only to use AI but to question its inevitability, refuse its extraction, and participate in collective decision-making about when AI belongs in education. This resonates with related critical work on the power-knowledge dimensions of AI literacy, the (im)possibility of AI literacy, and the principled teaching of AI education. For educators, it urges shifting from "ethical use" checklists toward deliberative community decision-making; for institutions, it raises questions about funding and infrastructural incentives that favor corporate interests over community benefit.

Limitations

As a perspective paper, it presents an argumentative, conceptual position rather than empirical evidence of effectiveness. The authors acknowledge practical challenges: financial constraints, sociopolitical pressures, and logistical difficulties in coordinating diverse groups; systemic inertia and educator compromises when schools resist critical perspectives; and the risk of co-opting community voices. They also note the paper's grounding in western, Anglophone academia and caution that they cannot claim expertise over the lived experiences of communities they do not directly belong to, and that applicability may be limited in regions with censorship or high institutional barriers.

Connected Concepts

Connected Articles

Citation

Li, S., & Mroziak, J. (2026). Reorienting Critical AI Literacy: A Community-Rooted Praxis of "Resisting AI". In Proceedings of the 2026 Conference for Research on Equitable and Sustained Participation in Engineering, Computing, and Technology (RESPECT 2026). ACM, Chicago, IL.