Every international school is writing AI policy in isolation. Without developmental science. Without a shared evidence base. Without a governance architecture that reflects how children's brains actually develop.

Most school AI policies reduce governance to academic integrity: how to detect plagiarism, when to declare tool use, and what constitutes cheating. These are secondary administrative concerns. Unrestricted cognitive offloading during critical neurodevelopmental windows is an environmental intervention. The issue is not whether an institution has an AI policy, but whether that policy is accountable to the biological development of the human prefrontal cortex.

The JAI Behavioural AI Governance Framework

A clinically grounded developmental architecture addressing four institutional constituencies: leadership, faculty, students, and parents.

• Ages 3–6 · Baseline Insulation

  Zero generative AI exposure. Strict preservation of sensorimotor development and early foundational cognition.

• Ages 7–10 · Observational AI Literacy

  Conceptual understanding and digital literacy without personal generative use. Full protection of unassisted synthesis and mechanical fluency.

• Ages 10–13 · Scaffolded Tutoring

  Narrow, hint-only interaction under strict educator supervision. Explicit restriction against autonomous text generation.

• Ages 13 and Above · Monitored Governance

  Supervised cognitive interaction within institutional data perimeters, paired with mandatory unassisted evaluation baselines.