ABSTRACT
Recent legislative developments in the European Union, the United States, and other jurisdictions have formalized the ‘right to repair’. A premise common to these legislative frameworks is that the object of repair is a tangible artifact completed at the time of sale, and that repair consists in returning that object to a predetermined reference state – what this Article terms the ‘static product model’. Contemporary AI-enabled systems, however, continuously evolve through ongoing interaction with users, data, and environments. When such systems deviate from expected behavior, users frequently lack the technical and legal means to ‘repair’ them. The root of the problem lies in the ontological understanding of technical objects presupposed by existing right-to-repair frameworks. This Article responds to this problem by drawing on the philosophy of technology of Gilbert Simondon. Simondon conceived of technical objects not as fixed entities but as processes that continue to individuate in relation to what he calls the ‘associated milieu’. From this perspective, repair is to be redefined not as a return to a prior state but as participation in a continuing process. Moreover, in Simondon’s account, the institutional exclusion of users from the formation of the associated milieu constitutes a form of technical alienation subject to normative criticism. On the basis of this theoretical reconstruction, this Article examines the institutional implications of a ‘right to repair as participation’ – reconsidering, in light of existing legal frameworks, a range of responses from the imposition of obligations regarding technical means such as API disclosure and interoperability guarantees to the institutionalization of procedural participation modeled on environmental impact assessment law.
Akasaka, Ryota, When Right to Repair Meets AI: Reinterpreting the Right to Repair through Individuation (March 17, 2026).
Leave a Reply