Drafting an updated AI use policy for our k-12 district (board adoption planned for august). current draft is rough. looking for district-level policy examples that have actually been adopted (not just proposals), particularly ones that handle the disclosure/disability accommodation tension well
Three district policies worth reviewing that handle the accommodation question well: 1) Boston Public Schools 2025 policy - distinguishes ‘general use’ from ‘accommodation use’ explicitly. Accommodation use doesnt require disclosure if documented in IEP or 504 plan. 2) Toronto District School Board 2026 policy - tiered by grade level (elementary, middle, high) with different defaults. Includes safe-harbor for student honest mistakes. 3) Fairfax County (VA) 2025 policy - has the cleanest disclosure language and is widely-cited by other districts as a template. Common patterns across the well-functioning policies: clear AI use taxonomy (correction tools vs generation tools), accommodation provisions tied to existing IEP/504 frameworks, escalation path for disputes that doesnt put detection software output as the final word, and explicit teacher training requirements before enforcement begins. The training piece is the part most districts skip and then regret. Happy to send the three policy docs - DM me.
Adjacent observation from corporate side: the policies that include training requirements get adopted smoothly. The ones that dont, get rolled back within a year because the implementation fails. Training is non-negotiable.
the accommodation provisions tied to IEP/504 is the framework that worked at my undergrad institution too. its an existing legal framework so theres less risk of inadvertent ADA violations. good model.
@ThomasLnrd DMing for the Boston, Toronto, and Fairfax policies. The taxonomy + accommodation framework is what our draft is missing. Updating before board adoption. Thank you for the comprehensive answer.