A.N.I.M.A.: An Experimental Pedagogical Model for Conscious Self-Narration in First-Cycle School

Learning to narrate oneself is a decisive competence in primary and lower-secondary school. Recounting a personal episode requires selecting events, ordering them, assigning them meaning and translating them into words: a formative practice that gives shape to experience, names emotions and builds awareness of the self (Bruner, 1991; Demetrio, 1996). Psychology has shown how identity itself takes the form of an internalised and evolving life story. Generative artificial intelligence, now within reach even of young students, is however most often encountered in a substitutive mode, with the machine being asked to write in the student’s place. Applied to autobiography, this mode delegates the very act that constitutes its core: a self-narrative generated by AI is, strictly speaking, no one’s narrative. A.N.I.M.A. proposes a reversal: AI does not write the self, but reads it. It works as an interpretive lens that makes recurrences, emotions and structures visible in texts students have already produced, returning a representation on which to reflect. The risk to be averted is cognitive offloading, which weakens the monitoring of one’s own learning (Fan et al., 2025). The model, evidence-informed in design, is consistent with the 2025 National Guidelines (D.M. 221/2025), which gradually replace those of 2012 by placing the person, identity and linguistic and reflective competences at the centre, and with the ministerial Guidelines on AI (D.M. 166/2025), which prescribe human responsibility, teacher supervision, protection of minors’ data and the critical use of these systems. Four principles underpin its design: the student’s authorial centrality, teacher mediation, AI as an object of inquiry rather than an anthropomorphised interlocutor, and metacognitive reflection (Flavell, 1979; Long & Magerko, 2020). The pathway unfolds in four phases whose Italian names form the acronym A.N.I.M.A. (meaning “soul” in Italian) which is retained unchanged in English: Autobiografia Narrativa (Narrative Autobiography), Indagine (Inquiry), Mediazione algoritmica (Algorithmic Mediation) e Autoconsapevolezza (Self-awareness). The following description uses the English translations of each phase name. In Narrative Autobiography, a human-only phase, students write a text about a significant experience within a protected space and without any judgement on form, with prompts differentiated between primary and lower-secondary levels. In Inquiry, students question the AI with analytical prompts to observe how the machine reads the text, prevailing emotions, recurring words and narrative structure, after the anonymisation of sensitive data. Algorithmic Mediation is the model’s centre of gravity: the student compares what they meant to say with what the AI returned, and the machine’s error becomes the occasion to reaffirm their own communicative intention and to understand that the system processes language probabilistically. In Self-awareness, finally, the text is reworked into an artefact (a revised version, a reflective journal, an emotion map, a comparison sheet), with the option to reject the AI’s interpretations with explicit justification. The expected advantages concern the safeguarding of authenticity, the development of metacognition, algorithmic literacy, metalinguistic reflection and personalisation. Critical issues remain to be managed: the privacy of the texts, the anthropomorphisation of the machine, the risk of dependence, algorithmic bias (Baker & Hawn, 2022) and the need for adequate teacher training. Assessment is formative and rests on a portfolio, a learning log and a four-dimension rubric: autobiographical awareness, metalinguistic reflection, critical capacity towards the output, and digital responsibility. This rubric aims to assess the process rather than the product. The contribution is theoretical and design-oriented and awaits empirical validation, ideally through a comparative design contrasting classes that practise traditional autobiographical narrative with classes that adopt the model. Its value lies not in producing more correct texts, but in forming students capable of distinguishing lived experience from algorithmic interpretation and of guarding that distance with a critical spirit, in a world populated by machines that produce plausible narratives yet lack any lived experience.