Trustworthy AI Governance in Cooperatives: A Multidimensional Framework for Financial Decision-Making
Dorleta Urrutia Oñate’s report develops a framework for governing AI in cooperatives that treats trust as something built, tested, and maintained through everyday institutional practice. Drawing on the Mondragón system, it argues that cooperatives already possess the conditions needed for trustworthy AI: democratic ownership, distributed accountability, and participatory oversight. The core contribution is a four-part model—technical, organizational, social, and institutional—that links model validation and explainability to governance structures, member participation, and regulatory alignment. Rather than relying on external auditors or new institutions, the paper shows how existing cooperative bodies—assemblies, councils, and federations—can absorb AI oversight. Trust, in this view, is not a principle but a process: sustained through documentation, education, contestability, and collective control.
Cite:
Oñate, Dorleta Urrutia. Trustworthy AI Governance in Cooperatives: A Multidimensional Framework for Financial Decision-Making. New York: Institute for the Cooperative Digital Economy, 2026.