Why your school's AI use policy should be three pages, not thirty.
Trustees love long policies. Teachers ignore them. Parents don't read them. The seven-template AI Policy Starter Pack Deshika ships with every engagement is three pages per template on purpose. Here is why short policy is better policy.
The seven templates cover what an Indian school actually needs to publish before April 2026. There is an Acceptable Use Policy for students, drawn from ASCD and Singapore MOE precedent. There is a Teacher AI Use Policy anchored on the NAESP and NASSP US principal-association guides, which names what teachers may use AI for and what they may not — final grades, IEP decisions, and student-record narratives without human review stay off the list. There is a Student-Facing Tool Policy, deliberately tool-agnostic so it doesn't end up looking like the vendor wrote the policy and sells the only tool the policy approves. There is a Parent Communication template, a DPDP-aligned Data Privacy and Child-Safety Statement, an AI in Assessment policy that frames "is this good work" as the question rather than "is this cheating," and an Incident Response procedure aligned to the school's existing crisis-comms structure.
The discipline is one thing. The length is another. Every one of these is three pages. The reason is operational, not aesthetic. A thirty-page policy in 2026 is a policy nobody on staff has read. A three-page policy is one the section coordinators have actually seen, can quote at parent meetings, and remember when a real incident lands on a Tuesday afternoon. The trustee can sign it without two weeks of redlining. The DPO can read it before the procurement decision. The class teacher can look at it in five minutes the morning before the first AI lesson.
The conflict-of-interest guard is non-negotiable. Policies stay tool-agnostic. They name categories of tools — "student-facing tutors," "teacher copilots," "plagiarism-detection" — never specific vendor names. Deshika's own name does not appear in any of the seven policies the school co-authors. The tutor approval is a separate decision gate, a separate document called the school's "authorised-tools list," maintained outside the policy itself. This is how the procurement reviewer in 2028 — checking whether the vendor sold a policy that recommends the vendor's own product — finds nothing to flag.
The Deshika engagement does the policy work as a two-day workshop on the school premises. Day one: open seven tabs on the projector, one per template. Walk the senior teachers, the IT lead, and the principal through each one. Vote on every contested clause. By end of day one, the student Acceptable Use Policy is drafted. By end of day two, all seven are signed. The principal's name goes on them. Deshika's name does not. The policies stay with the school whether the school stays with Deshika or not.
The long-game move is the open-source step. After Deshika has co-authored twenty-five real school AI policies — Q3 2026 target — the templates open-source under Creative Commons. The brand position cements as "the company that wrote India's AI-policy library," and the procurement reviewer finds public templates that any vendor can reference. That is what turns policy-authorship into a moat instead of a lock-in.
The policy is who the school is. The tool is what the school uses. The first one outlasts every vendor decision. That is why it should fit on three pages, and why the school's name should be the only name on it.