ETICA SOFT PER UN'IA RESPONSABILE GOVERNANCE NEGLI AMBIENTI EDUCATIVI DELLA PRIMA INFANZIA

This thesis investigates how soft ethics can support responsible AI governance in early childhood education (ECE), where digital tools increasingly shape learning, relationships, and children’s emerging moral agency. It asks how AI ethics can safeguard the image of the child as a creative, relational and capable person, and how ethical instruments can be embedded in governance so that AI serves children’s flourishing rather than reducing their lives to data and prediction. The introductory chapter situates AI governance within contemporary debates on freedom, sustainability and global regulatory regimes, and defines core concepts such as agency, flourishing and soft ethics. Chapter 1 explores the image of the child in the age of AI, tracing the development of moral agency from early childhood through insights from moral psychology, neuroscience and Thomistic philosophical anthropology, and argues that AI ethics in ECE must ultimately protect this rich, developmental image. Chapter 2 maps selected Western as well as non-Western ethical traditions, alongside meta-ethical frameworks, and proposes an intercultural method for translating diverse moral languages into a shared, yet plural, basis for techno-moral judgment. Building on this foundation, Chapter 3 analyses how AI reshapes early childhood education through stakeholder dynamics, digital identities, and concrete ethical risks such as surveillance, bias and anthropomorphisation. It shows why existing AI-ethics frameworks lack a stable normative ground and points to the need for soft-ethical governance. Chapter 4 then operationalises soft ethics by proposing a dual governance architecture that combines institutional deliberation—via AI Ethics Committees—with technical AI Ethics Guardrails embedded in agentic systems, together forming a moral infrastructure for child-centred value alignment. Chapter 5 instantiates the first pillar of this model through a policy-oriented proposal for an AI Ethics Committee within Italy’s 0–6 education governance, grounded in Reggio Emilia, Montessori and Catholic pedagogical traditions 2 and positioned at the meso-level as a participatory, anticipatory mechanism of oversight. Together, these chapters advance the novel argument that soft ethics must be operationalised as a practical governance resource—one that enables AI in early childhood education to remain developmentally appropriate, relationally attentive, and fundamentally oriented to the long-term flourishing of children.