Microsoft Sets New AI Privacy Rules for US Schools

Microsoft has established a new agreement containing ten safety and privacy principles for artificial intelligence in classrooms. Developed alongside the American Federation of Teachers (AFT), which is the second-largest teachers union in the United States, and its New York City affiliate, the United Federation of Teachers (UFT), the terms are contractually enforceable. Starting in November, school districts can choose to integrate these specific terms into new or existing contracts with Microsoft without needing to renegotiate their entire agreements.

Under the new guidelines, Microsoft commits to not training its AI models on student or educator data. The company will also limit its initial data collection, explain how its tools function to families in plain language, ban AI companions, and mandate human oversight for high-risk decisions. This policy shift follows recent actions by school districts in New York City and Los Angeles, both of which enacted one-year bans on student-facing AI tools to allow time to evaluate safety guardrails.

The AFT has actively pushed for guardrails while preparing educators for technological shifts. AFT President Randi Weingarten previously called for a ban on screens before third grade and student-facing AI before middle school. However, the union is also collaborating with tech firms, recently launching a $23 million AI training hub for educators funded by Anthropic, Microsoft, and OpenAI. Weingarten praised the new Microsoft agreement as legally binding, noting that "no one else, including the federal government, has stepped up" to establish these protections.

For school administrators and educational technology coordinators, this agreement provides a ready-made legal framework to protect student privacy without halting AI integration. Instead of navigating complex, individual negotiations with tech giants, districts can easily adopt these standardized protections. For developers and practitioners in the educational AI space, Microsoft's concession sets a new industry benchmark, signaling that future classroom tools must prioritize strict data boundaries and human-in-the-loop decision-making to gain widespread institutional trust.