Derecho a la dificultad y gobernanza democrática de la inteligencia artificial educativa

The objective is to determine which operations may be delegated to generative artificial intelligence without displacing the experiences through which capacities are developed and learners become accountable to others. The methodology is based on a philosophical-normative reconstruction and draws on empirical evidence as a means of examining the argument rather than as an independent foundation for the thesis. The approach conceives of education as the development and appropriation of capacities—not merely the production of outputs—and examines delegation in terms of future autonomy, epistemic agency, relational responsibility, and public accountability. The analysis develops two concepts: the “right to formative difficulty”, a prima facie institutional claim against the unjustified substitution of operations essential to capacity development; and “governable dependence”, understood as a form of mediation that is intelligible, contestable, and compatible with retaining authority over educational aims and evaluative criteria. Both concepts are examined through three justificatory tests. The article concludes that AI may legitimately participate in distributed cognitive practices, but should neither replace, without justification, the experiences through which capacities are developed nor independently determine educational decisions affecting individuals. The democratisation of AI in education requires institutional guarantees of participation, contestation, and the possibility of withdrawing from or replacing these systems.