Microsoft and the American Federation of Teachers agreed a privacy standard for AI in schools. It sets 10 protections the union says are legally enforceable. Microsoft will not use student or teacher data to train or improve AI models. It carved out narrow safety and security exceptions. The standard bars tracking of students and requires human oversight of school decisions. Microsoft must also explain in plain language how its tools work. Breaches would expose the company to contract liability. Microsoft must publish reports, certify annually and fix security issues to deadlines. Districts can add the terms to existing contracts from 1 November. The union says it is in talks with OpenAI and Anthropic to sign the same terms.
What changed
No federal or state rules governed AI use in US schools, and districts negotiated protections alone.
What it unlocks
US school districts can attach the privacy terms to existing Microsoft contracts without renegotiating.
- 10 enforceable protections
- available to districts from 1 Nov 2026
- $23 million backing the AI academy
- 200,000 UFT members
What you need to act on it
- a US school district
- an existing or new Microsoft customer agreement
Sources