🔍 Read the full analysis: What’s Actually Required For Human Oversight Of AI In Schools? on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Microsoft says that from November 1, 2026, U.S. school districts can use a contract framework restricting how education-focused AI handles student and educator data. It bars model training on that data, tracking and companion-style features, and requires meaningful human review of decisions—but it is a contract, not a law, and its scope and adoption remain important limits.
Microsoft says U.S. school districts can use a new contract framework from November 1 that limits how education-focused AI may handle student and educator data, track users, imitate relationships or make decisions. The 31-page agreement requires meaningful human review for covered decisions, but it is a contract districts can adopt—not a nationwide law automatically binding every school or AI provider.
The agreement, called the National AI Safety & Privacy Standard, was announced September 9 by the American Federation of Teachers (AFT), the United Federation of Teachers (UFT) and Microsoft. Its formal structure is a memorandum between the National Academy for AI Instruction, an educator-led training body announced by the AFT in 2025, and Microsoft. The document includes empty signature spaces for other AI providers. Districts can incorporate its terms into their own Microsoft contracts.
For covered products, providers may not use prompts, answers, uploaded files, behavioral signals or metadata from students, teachers and administrators to train, fine-tune, benchmark or otherwise improve AI models. The ban also applies to de-identified data and continues after a contract ends. A narrow exception permits processing to identify specified safety risks, including self-harm, abuse material, grooming, threats, bullying and security attacks; that information may not then be used for general model improvement.
The agreement also bars specified forms of tracking, including continuous passive attention monitoring, long-term profiling and biometric or emotional inference, subject to an exception requiring a school customer’s explicit written approval. It prohibits companion-style features designed to foster emotional attachment or dependency. Covered systems may not make decisions about students or teachers without meaningful human review or prior approval. Student-facing agents that act outside a provider’s systems must be off by default unless an administrator enables them.
Infographic 01 · The lines drawn this autumn
Seven weeks, one direction: a human stays in charge
From the school contract to state laws: what was signed when, and what the classroom standard actually requires.
- 2 Feb 2025EU: emotion inference banned at work and schoolAI Act Art. 5(1)(f), except medical or safety reasons
- 9 Sep 2026AFT, UFT and Microsoft announce the school standardA contract districts can add to their Microsoft agreements
- 30 Sep 2026California signs AB 1883, SB 947, SB 951, AB 1979Emotion AI at work, robot bosses, AI layoff notices, AI in clinics
- 28 Oct 2026New York signs its chatbot law for minorsS.9051B/A.10379C: no unsafe companion features for under-18s
- 1 Nov 2026Microsoft makes the standard available to all US districtsPer Microsoft’s fact sheet; districts “can add them”
- 1 Jan 2027Most California rules take effectAB 1883, AB 1979, SB 951
- 1 Jul 2027SB 947 becomes operativeHuman corroboration before AI-driven discipline or firing
What Microsoft’s classroom AI may not do
- Learn from students or teachers. No training on their data; de-identification is “not a loophole”.
- Track them. No keystroke logging, attention monitoring, long-term profiling or biometrics.
- Befriend them. No companion features designed to foster emotional attachment.
- Decide alone. No decisions without meaningful human review or prior approval.
Where it is narrower
- Scope: education products only. General-purpose productivity and workplace tools are excluded by definition.
- Opt-in or automatic: Microsoft says districts “can add” the terms; the agreement says “upon request within 90 days”.
- Exceptions: a narrow safety exception for training; the no-tracking list allows written approval by the school.
“Anything less than legally enforceable provisions is simply a wish list.”Randi Weingarten, AFT President, 9 September 2026
Sources: AFT/UFT/Microsoft release, 9 Sep 2026 · Standard MOA v1.0 · Microsoft fact sheet · Governor of California, 30 Sep 2026 · Governor of New York, 28 Oct 2026 · EU AI Act, Art. 5(1)(f). Effective dates for California follow the chaptered texts and the state’s default start date for regular-session bills.
Powered by Thorsten Meyer AI
Contract Terms Put Review in District Hands
The agreement gives schools specific terms to put into procurement contracts rather than relying only on a provider’s general privacy statements. The document says a district may end a contract and seek damages if its terms are breached, making enforcement depend on the agreements districts sign and the remedies available under them.
For educators and families, the human-review language addresses a practical concern: AI can inform classroom work, but a covered tool should not make consequential decisions about a student or teacher without a person reviewing or approving them. Limits on emotional inference and companion-style features also set boundaries around the kinds of student information and interactions the agreement permits.
The protections are not universal. The agreement covers AI products designed and marketed for education; general-purpose productivity and workplace tools are expressly excluded. Districts must also determine whether and how to adopt the terms. The source material describes the agreement’s reach differently in places, so availability should not be confused with automatic coverage of every school system or product.
Products Worth Considering
Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Black
Apple Watch Series 12 GPS 46mm Smartwatch
Logitech MX Keys S Wireless Keyboard, Graphite
As an affiliate, we earn on qualifying purchases.
A Contract, Not a National Law
The education agreement is one part of a wider set of developments described in the source material. On September 30, California Governor Gavin Newsom signed four measures addressing AI in workplaces and health care. The source says AB 1883 restricts AI emotion reading and collection of neural data from workers; SB 947 bars discipline or firing decided by an algorithm alone; AB 1979 limits AI from performing work reserved by law for licensed health professionals; and SB 951 requires layoff notices to identify when AI replaced jobs. Most are scheduled to take effect January 1, 2027, while SB 947 is scheduled for July 1, 2027.
On October 28, New York Governor Kathy Hochul signed a law addressing certain chatbot companion features for minors and unverified users. The source describes restrictions on chatbots that pretend to be human or foster emotional dependency. These state laws and the school agreement use different legal mechanisms: state statutes impose legal requirements within their scope, while the school standard works through contracts districts choose to sign.
The agreement’s stated operational requirements include reporting breaches within 72 hours, deleting data from active systems within 180 days, and keeping data in the EU or North America. Those terms sit alongside the substantive restrictions, but they do not expand the agreement to products or districts outside its stated contractual scope.
“Anything less than legally enforceable provisions is simply a wish list.”
— AFT President Randi Weingarten
Adoption and Coverage Remain Open
It is not yet clear how many districts will adopt the agreement or how quickly providers other than Microsoft will sign it. Microsoft’s announcement says protections are available to every U.S. district from November 1, but the agreement operates through district contracts. The source also describes its application differently in different places, leaving open whether districts must request the terms or whether some Microsoft arrangements include them by default.
The exact boundary between education-focused products covered by the agreement and general-purpose tools excluded from it may matter for schools using the same AI provider across classroom, administrative and workplace functions. The document’s school-approved exception to some tracking restrictions also means the precise protections can depend on written district decisions. The available source material does not establish how compliance will be audited or how disputes will be resolved in practice.
Districts Decide What to Sign
School districts will need to review their AI procurement and Microsoft contracts to determine whether to incorporate the terms and which products they cover. They will also need to decide how to handle any written approvals for tracking exceptions, who reviews AI-assisted decisions, and how staff can report possible breaches.
Providers may choose whether to sign the agreement, while districts and families will be able to watch for evidence of adoption and compliance. The next practical test is whether the contractual safeguards appear in district agreements and are followed in day-to-day use. Separately, California’s scheduled effective dates in 2027 will bring its workplace and health-care measures into force according to the timetable described in the source.
Key Questions
Does the agreement automatically apply to every U.S. school?
No. Microsoft says the standard is available to every U.S. school district from November 1, 2026, but it works through contracts. District adoption and the precise way the terms enter a district’s agreement remain material.
What does the agreement say about student data?
For covered education products, student and educator data—including prompts, uploaded files and metadata—may not be used to train or improve AI models. A limited safety-processing exception is allowed for specified risks, but the data cannot then be used for general model improvement.
Does human oversight mean AI cannot help make school decisions?
No. The agreement allows AI tools to support educators, but says covered systems cannot decide matters about students or teachers without meaningful human review or prior approval. The source also says the educator training body will instruct teachers not to use AI for automated grading without review, discipline, placement or emotional assessment.
Which AI products are covered?
The stated scope is AI products designed and marketed for education. General-purpose productivity and workplace tools are expressly excluded, so a school will need to check which products and contracts fall within the terms.
Is this agreement the same as a law?
No. It is a contract framework that districts can include in agreements with Microsoft. Its enforcement depends on contractual adoption and terms; it does not by itself impose a nationwide statutory rule on every school or AI provider.
Source: ThorstenMeyerAI.com
