Microsoft Agrees to Binding AI Privacy Rules for Schools

Microsoft has signed a legally binding agreement with the American Federation of Teachers establishing 10 privacy and safety protections for artificial intelligence use in schools, marking the first such enforceable standard in U.S. education.

Under the agreement, Microsoft commits not to use student or teacher data to train or improve AI models except in narrow safety and security situations. The company also agrees not to track students and requires human oversight for any AI-driven decisions in educational settings.

The standard becomes available to every U.S. school district starting November 1. Districts can add these protections to existing Microsoft contracts without renegotiation, according to details first reported by POLITICO.

Why it matters

This agreement represents a private-sector response to the absence of federal AI regulation in schools. With Congress unable to pass legislation establishing AI guardrails for education, unions and tech companies are creating their own frameworks. The legally enforceable nature of these provisions—backed by breach-of-contract liability—gives them more weight than voluntary guidelines, potentially setting a template for the industry.

What the agreement covers

The 10-point standard requires Microsoft to provide educators and parents with clear, plain-language explanations of how AI tools function in classrooms, what controls schools have, and what safeguards protect students. Microsoft Vice Chair and President Brad Smith said the standard "sets a high bar for child privacy and AI safety."

AFT President Randi Weingarten described the agreement as having "real teeth," noting that Microsoft faces contract breach liability for violations. "We can get angrier and angrier, or we can act decisively," Weingarten said. "Anything less than legally enforceable provisions is simply a wish list."

The United Federation of Teachers, New York City's teachers union with 200,000 members, partnered on the agreement through AFT's National Academy for AI Instruction.

Broader context

The announcement follows recent AI restrictions in major school systems. New York City, the nation's largest district, announced a one-year AI moratorium for students in grades 2K through eighth earlier this month. Los Angeles Unified School District went further with a complete AI ban for all students.

AFT is in active discussions with OpenAI and Anthropic to adopt similar privacy standards, according to a union spokesperson. Last year, AFT launched the National Academy for AI Instruction with $23 million in support from Microsoft, Anthropic, and OpenAI—a partnership that drew criticism from some teachers concerned about tech company influence over classroom AI use.

The agreement builds on AFT's separate 10-point plan from May calling for screen bans for prekindergarten through second grade and restrictions on student-facing AI in elementary schools, with exceptions for students with disabilities requiring assistive technology.

Details of the agreement were first reported by POLITICO.