Conference Program

The growing integration of artificial intelligence in educational contexts often promotes efficiency-oriented models of learning based on prediction, standardization, and the delegation of judgment, where AI is approached as an immediate source of answers rather than as an object of inquiry. In this scenario, this contribution reframes AI education through the perspective of convivial and non-directive pedagogies, understood as practices capable of sustaining students’ autonomy and the plurality of interpretative processes.

This contribution draws on a situated educational experience in upper secondary education (grades 10–12) and is based on participant observation of a cycle of workshops specifically designed to critically explore the role of AI in learning processes. In this context, AI is not introduced as a tool for optimizing performance or providing correct solutions, but as an object of shared exploration and as a device for the dialogic co-construction of knowledge. The activities are designed to suspend the logic of the single correct answer and immediate validation, and to privilege the comparison of different interpretations, collective argumentation, and the negotiation of meaning within the classroom.

Within this setting, AI’s epistemic authority is deliberately decentered and interaction with generative systems becomes a participatory space in which knowledge is not received but constructed through dialogic practices. From this perspective, AI is framed neither as a neutral device nor as a performance-support tool, but as an element that renders visible the tensions between standardization and autonomy and enables a democratic reconfiguration of the educational environment.

The theoretical framework is grounded in the traditions of convivial and non-directive pedagogies. Illich’s notion of convivial tools allows AI to be interpreted as a technology that can either sustain or constrain autonomy depending on the educational practices in which it is embedded, while Freire’s pedagogy of autonomy informs the design of learning environments based on dialogic responsibility, participation, and the co-construction of knowledge.

The observations collected indicate that practices of this kind foster a use of AI that is not oriented toward delegation, but toward the production of spaces for plural reasoning in which students are called to take a position, engage with different perspectives, and actively participate in knowledge construction. In this light, AI education does not coincide with the acquisition of technical skills, but with the design of convivial contexts capable of sustaining autonomy, reflective judgment, and democratic participation.

This contribution engages with the panel by showing how the principles of non-directive pedagogy can be translated into concrete educational practices and how AI can be reconfigured, within such practices, from an apparatus of normalization into a convivial partner for democratic learning.