Same basis, not a lower bar: design of SameBasis, a teacher-controlled AI decision-support prototype for reasonable adjustments in Australian schools

In 2025, 27.0% of Australian school students received an educational adjustment due to disability. Under the Disability Standards for Education 2005, each adjustment must let the student take part on the same basis as their peers, yet the 2020 Review of the Standards found that many educators struggle to know how. This design report presents SameBasis, a web-based prototype in which generative AI drafts one specific, reasoned adjustment for a named student in a named activity, covering UDL, cultural and trauma-informed reasoning. The teacher always decides, implemented adjustments become NCCD-ready evidence, and an equity monitor makes AI use visible to the school. The paper sets out six design principles, maps features to obligations, states limitations plainly (synthetic data only, not yet evaluated, not a compliance tool) and proposes a staged evaluation plan. Stratedge Working Paper WP 2026-01.