🔍 Read the full analysis: What Schools Need To Know About Human Oversight In AI Rules on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Microsoft says that starting Nov. 1, 2026, every U.S. school district can access a contract setting limits on AI products designed and marketed for education. The agreement bars training on student and educator data, restricts tracking and companion-style features, and requires meaningful human review for decisions. It is a contract, not a law, and its coverage and application have limits.
Microsoft says that starting Nov. 1, 2026, every U.S. school district can access a contract requiring human review of decisions made with covered classroom AI. The agreement also restricts training on student and educator data, tracking and companion-style features, but applies to AI products designed and marketed for education and is not a federal law.
The 31-page Memorandum of Agreement was announced Sept. 9 by Microsoft and the teachers’ unions AFT and UFT. It is between Microsoft and the National Academy for AI Instruction, an educator-led training body announced by AFT in 2025. The document includes empty signature slots for other AI providers. Microsoft says it is making the protections available to every U.S. school district from Nov. 1.
The agreement bars covered providers from using prompts, answers, uploaded files, behavioral signals and metadata from students, teachers and administrators to train, fine-tune, update, benchmark or otherwise improve AI models. It says de-identification does not create an exception. A narrow exception permits processing data to detect risks such as self-harm, abuse material, grooming, threats, bullying or security attacks; that data cannot then be used for general model improvement.
Other terms prohibit specified forms of tracking, including continuous passive attention monitoring, long-term profiling and biometric data collection, including emotional inference. The text allows some listed tracking exceptions with explicit written approval from the school customer. It also bars 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. The agreement sets additional requirements for incident reporting and deletion of data.
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
How Contract Terms Reach Classrooms
The agreement gives districts a potential way to put limits on covered AI products through procurement contracts rather than waiting for a new law. According to the source material, districts can incorporate the terms into their own Microsoft contracts; breach could allow a district to end the contract and seek damages. That makes the terms relevant to school purchasing decisions, though the precise process for adoption matters.
For educators and families, the human-review provision draws a line between using AI to assist and allowing it to make consequential decisions on its own. The restrictions on data use and monitoring also address information students generate during learning. These are contractual protections, not a blanket ban on AI in schools, and the education-specific scope means the agreement does not cover every tool a district might use.
The announcement sits alongside government rules in other settings, but those should not be confused with the school contract. California enacted laws concerning workplace emotion analysis, algorithmic employment decisions and health care. New York signed a law addressing certain chatbot features for children. Together, the developments point to attention on human responsibility and limits on sensitive data, while each has a different legal basis and scope.
Products Worth Considering
FLEXISPOT EN1-PRO Electric Standing Desk, 79 x 32, Black Walnut
Logitech MX Keys S Wireless Keyboard, Graphite
Canon IVY 2 Mini Photo Printer Bundle, Pure White, with 110 ZINK Prints and Protective Case
As an affiliate, we earn on qualifying purchases.
Three Separate Rules, Different Reach
The school agreement is a memorandum between organizations, not a statute or a general Microsoft policy page. The source says districts can write its terms into Microsoft contracts and that other providers have signature slots, but it does not establish that all providers have signed or that every district is automatically covered.
California Gov. Gavin Newsom signed four measures on Sept. 30, 2026, according to the source. AB 1883 addresses AI that reads workers’ emotions or collects neural data; SB 947 bars firing or discipline decided by an algorithm alone; AB 1979 limits AI from performing health-care functions reserved for licensed professionals; and SB 951 requires layoff notices to disclose when AI replaced jobs. Most take effect Jan. 1, 2027, while SB 947 takes effect July 1, 2027.
New York Gov. Kathy Hochul signed a law on Oct. 28 addressing certain chatbot companion features for minors and unverified users. The source describes restrictions on unsafe features, including chatbots that pretend to be human or encourage emotional dependency. These state laws are distinct from Microsoft’s district contract and do not, by themselves, establish rules for every school AI product nationwide.
“Anything less than legally enforceable provisions is simply a wish list.”
— AFT President Randi Weingarten
Who Is Covered by the Agreement
The source material gives differing descriptions of whether the terms apply automatically or only when a district asks to include them. Microsoft says the contract is available to every district, while the agreement’s described mechanism relies on districts writing its terms into their contracts. The practical effect may depend on what each district signs.
It is also unclear how many districts will adopt the terms, which other AI providers will sign the memorandum, and how disputes over compliance would be handled. The agreement’s scope excludes general-purpose productivity and workplace tools, according to the source, so a product’s intended design and marketing may affect whether the protections apply. The listed school-approved exceptions to some tracking prohibitions also leave implementation details to districts.
The source does not provide independent legal analysis or the full text of the California and New York laws. Their exact application will depend on statutory language and implementation. Readers should distinguish the reported summary of those laws from the specific terms of the school contract.
Districts Decide What to Sign
School districts can review the memorandum and decide whether to incorporate its terms into their Microsoft contracts. The next practical milestones are district procurement decisions, any further provider signatures, and the handling of complaints or reported breaches. The agreement sets a 72-hour breach-reporting requirement and says deleted data must be removed from active systems within 180 days.
Districts and educators will also need to identify which tools are designed for education and whether the relevant terms are included in the contracts they use. The source recommends reviewing features associated with sentiment, mood and engagement, as well as decisions involving discipline or dismissal. Those are suggested checks, not a statement that every such feature is prohibited in every product or setting.
California’s measures are scheduled to take effect in 2027, on different dates, while New York’s newly signed law will require its own implementation. As those rules take effect and districts make procurement choices, the open question is how the written protections translate into day-to-day oversight and enforcement.
Key Questions
Is Microsoft’s school AI agreement a law?
No. It is a contractual memorandum between Microsoft and the National Academy for AI Instruction. Districts can incorporate its terms into their Microsoft contracts; it is not a law applying automatically to all schools.
What does the agreement require for decisions about students?
Covered AI systems may not make decisions about students or teachers without meaningful human review or prior approval. The agreement also says AI should support, not replace, human judgment.
Can student data be used to train AI models?
The agreement bars covered providers from using student and educator data to train or improve AI models. It describes a narrow safety-processing exception, but says data processed under it cannot then be used for general model improvement.
Does the agreement cover every AI tool a school uses?
No. The source describes the agreement as covering AI products designed and marketed for education. General-purpose productivity and workplace tools are expressly excluded, and district adoption of the contract terms is a separate issue.
When do the related state laws take effect?
According to the source, most of California’s listed laws take effect Jan. 1, 2027; SB 947 takes effect July 1, 2027. New York signed its chatbot law on Oct. 28, 2026, but the source does not specify its effective date.
Source: ThorstenMeyerAI.com
