Beyond Homework Battles: AI‑Assisted Learning Guidance and Parent–Child Relationships in Hong Kong SEN Families

Abstract

Hong Kong’s exam oriented education system and limited provision for students with special educational needs (SEN) place substantial strain on parents, and school assigned tasks for home completion can further exacerbate intergenerational conflict. This paper examines how generative AI might be used as a co mentoring tool in the home to reduce this strain, support learning, and strengthen relationships between parents and their SEN children in Primary 3–4. Drawing on evidence that Hong Kong SEN parents experience elevated stress, role overload, and feelings of incompetence, it argues that these pressures undermine communication and emotional attunement during homework and skills practice. The paper then considers recent developments in AI mediated personalisation—including the adaptation of task difficulty, the generation of interest based materials, and the production of social stories or visual supports tailored to children’s SEN profiles—and discusses how these affordances could be harnessed by parents. Building on this rationale, a constructivist framework for AI assisted co learning is proposed, in which parents are trained to co design AI supported activities and to shift from directive, correction focused teaching towards more collaborative, exploratory problem solving and positive feedback cycles. Finally, the paper outlines a mixed methods study to investigate changes in parent–child communication, stress, and perceived closeness when families adopt structured, AI supported home co learning routines. It is anticipated that this work will generate design principles for parent training programmes using AI tools that prioritise relational quality alongside academic outcomes.