What changed
Microsoft's AI Safety and Privacy Standard for schools took effect November 1, 2026. Any US school district can now add it to a Microsoft agreement with no renegotiation required. Once added, the district can cancel the contract and seek damages if Microsoft breaks the rules.
The standard permanently bans Microsoft from training AI models on student or educator data, with one narrow safety exception. It also bans selling student data, using it for ads, tracking students, and AI chatbots that act like a friend. High-stakes decisions require a person, not just AI. Breaches must be reported within 72 hours, and violations must be fixed or AI use paused.
The standard was negotiated by the American Federation of Teachers and took effect after New York City and Los Angeles school districts paused AI use over privacy concerns. OpenAI and Anthropic have expressed interest in signing similar agreements.
What you can do with it now
If your team works with schools, clinics, or any organization handling sensitive data, write the same demand into your vendor contracts. Not as a promise on a webpage, but as a term your lawyer can enforce. List what the vendor cannot do with your data: train models, sell it, track your clients, make decisions without you.
If you buy AI tools for your own team, check your agreement today. Does it promise your data won't train their next model? Can they change that promise by editing a webpage? If yes, ask for a contract amendment before you hand over another invoice or client file.
What to keep in your own hands
A contract is only as strong as your ability to know it was broken. Keep a list of what data you share, when, and with which tool. If you suspect a breach, you need records to prove it. And remember: a company can sign any standard and still go bankrupt, get acquired, or decide enforcement is too expensive to pursue. A good contract buys you the right to walk away. Make sure you can actually leave and take your data with you.