The AI-Based Identity Support Framework: Identity as an Architectural Design Variable in Refugee Education

More than 43 million children are currently displaced globally, yet the artificial intelligence (AI) systems deployed in their education are rarely designed with cultural identity, linguistic continuity, or belonging as architectural concerns. Existing frameworks treat identity as a contextual consideration surrounding educational technology; this paper argues that identity should be treated as a design variable within AI system architecture itself. The paper develops the AI-Based Identity Support Framework (AISF), a conceptual model constructed through interdisciplinary synthesis across bodies of literature including cultural bereavement theory, narrative identity, ethical AI design, and multilingual AI systems, and informed by the author’s practitioner engagement with displaced communities (2018–2023). The AISF specifies three interlocking domains (cultural anchoring, emotional co-construction, and adaptive belonging), each defined operationally and accompanied by diagnostic indicators, design principles, and a preliminary assessment rubric. The framework is differentiated from trauma-informed pedagogy, culturally sustaining pedagogy, and principle-based AI ethics as a system-architectural rather than pedagogical construct. The paper engages critically with risks internal to identity-supportive AI architectures, including essentialization, surveillance, and algorithmic identity-fixing, and addresses low-resource deployment constraints. The AISF is intended to provide a structured diagnostic instrument for designers, evaluators, and policymakers, aligned with Sustainable Development Goals 4.7, 10, and 16.