Microsoft and teachers unions set enforceable AI privacy standard for schools

US school districts will be able to add legally enforceable AI privacy and safety protections to their Microsoft contracts under a new agreement negotiated with leaders of the American Federation of Teachers and United Federation of Teachers.

The 30-page standard bars student and educator data from being sold, used for advertising or used to train AI models, apart from limited safety-related exceptions. It also prohibits student tracking and AI companion chatbots, and requires human oversight of AI decisions. Families must receive plain-language information about how school AI tools work, while districts retain control over data collection, storage, deletion and new product features.

The agreement adds practical enforcement measures. Microsoft will face audit and certification requirements, and data breaches must be reported within 72 hours and addressed. Districts that incorporate the provisions can cancel contracts and seek damages when rules are broken.

The standard does not endorse or ban classroom AI. Union leaders continue to support New York City’s temporary restrictions on student-facing AI and are seeking similar commitments from other developers. Important questions also remain: de-identified usage data may still be used to improve products, and research suggests anonymized information can sometimes be linked back to individuals. Even so, the agreement replaces broad promises with contractual obligations that schools can enforce.