🔍 Read the full analysis: How Microsoft’s School AI Pact And California’s Laws Keep Humans Involved on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Microsoft says that starting November 1, 2026, U.S. school districts can use a contract setting limits on AI products designed and marketed for education. Separately, California has signed laws restricting certain workplace and health-care uses of AI, most taking effect in 2027. The measures are not a single legal framework, and the school contract’s coverage and application depend on its terms and district procurement.
Microsoft says U.S. school districts can begin using a new AI contract on November 1, setting limits on how covered education products handle student data, monitor children and make decisions. The agreement is a contract template, not a nationwide law; its effect depends on whether districts incorporate its terms into their agreements. It comes as California has signed laws restricting specified workplace and health-care uses of AI, with most provisions due to take effect in 2027.
The 31-page Memorandum of Agreement was announced on September 9 by the American Federation of Teachers (AFT), the United Federation of Teachers (UFT) and Microsoft. It is between Microsoft and the educator-led National Academy for AI Instruction, with signature spaces for other providers. Microsoft says it is making the protections available to every U.S. school district from November 1. The source material describes districts as able to put the terms into their own Microsoft contracts; it does not establish that every district is automatically covered.
For covered education AI products, the agreement bars using student, teacher or administrator data to train, fine-tune, update, benchmark or otherwise improve AI models. The listed data includes prompts, responses, uploaded files, behavioral signals and metadata, and the text says de-identification does not create an exception. A narrow exception permits processing for specified safety purposes, including detecting self-harm risk, abuse material, grooming, threats, bullying and security attacks; that data cannot then be used for general model improvement.
The terms also restrict location and behavioral tracking, keystroke logging, continuous passive attention monitoring, long-term profiling and biometric processing, including emotional inference. They prohibit companion-style features intended to foster emotional attachment or dependency, and require meaningful human review or prior approval for decisions about students or teachers. The agreement allows some listed monitoring exceptions with a school customer’s explicit written approval. It also requires breach reporting within 72 hours and deletion from active systems within 180 days, according to the supplied account of the agreement.
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
Where Human Review Is Required
The measures speak to a practical question facing schools, employers and families: which decisions should remain with people when software can assess behavior, recommend action or interact with children. In schools, contract terms can give districts a basis to challenge a provider that violates agreed rules, rather than relying only on voluntary product promises. The source says a district may end its contract and seek damages for a breach, although the result would depend on the contract and circumstances.
California’s laws use a different route: state requirements rather than a district’s procurement terms. Together, the developments signal that human review, limits on sensitive data and restrictions on emotionally manipulative AI are becoming concrete issues in purchasing and compliance. They do not ban AI generally, and they do not establish that every automated recommendation is prohibited. Organizations will need to distinguish covered systems and decisions from uses outside the rules.
For educators and parents, the central distinction is between AI assisting with a task and AI being allowed to make consequential judgments about a person. For companies selling into schools, the agreement may affect product design, data handling and contract negotiations. For workers and health-care providers in California, the statutory rules may shape how employers and organizations use automated systems once the relevant provisions take effect.
Products Worth Considering
Sonos Era 100 SL Compact Microphone-Free Speaker – Black
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
Samsung 27-Inch Odyssey QD-OLED G8 4K 240Hz Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Three Routes to AI Limits
The school standard is a negotiated agreement, not legislation. Its scope is described as education products designed and marketed for that purpose. General-purpose productivity and workplace tools are expressly excluded in the supplied material, so the terms should not be read as covering all Microsoft AI products used by anyone connected to a school. The agreement also instructs educators not to rely on AI alone for automated grading, discipline, placement or emotional assessment.
California Governor Gavin Newsom signed four measures on September 30, according to the source. AB 1883 addresses AI that reads workers’ emotions and collection of neural data. SB 947 bars discipline or dismissal decided by an algorithm alone. AB 1979 limits AI in health care from performing work reserved by law for licensed professionals. SB 951 requires layoff notices to say when AI has replaced jobs. Most are scheduled to take effect on January 1, 2027; SB 947 is scheduled for July 1, 2027.
New York is another part of the emerging picture, but not a California law or part of Microsoft’s school agreement. The source says Governor Kathy Hochul signed a law on October 28 restricting specified “unsafe” chatbot companion features for minors and unverified users, including systems that pretend to be human or foster emotional dependency. The details provided do not include the law’s full scope or implementation schedule.
Coverage Depends on Contracts
It is not clear from the supplied material whether every district must affirmatively request or sign onto the terms, or how Microsoft will apply them across existing and new contracts. The account says the protections are available to every district, while also explaining that districts can incorporate them into their own contracts. Those descriptions do not establish automatic coverage.
The agreement’s practical reach is also limited by its stated scope: education products designed and marketed for education, with general-purpose productivity and workplace tools excluded. The full text, how the school-approved exceptions will be administered and how disputes would be resolved are not set out here. The source also does not provide full statutory text or implementation guidance for the California and New York laws, so their precise application should be checked against the laws and official guidance.
District Procurement and 2027 Rules
School districts can decide whether to incorporate the agreement into Microsoft contracts and how to handle any permitted monitoring exceptions. Vendors and district procurement teams will need to clarify which products are covered, what data is processed and how human review works. The agreement’s specified breach-reporting and deletion deadlines would matter if a covered provider violates the terms.
California’s next key dates are January 1, 2027, for most of the listed laws and July 1, 2027, for SB 947, according to the source. Employers, health-care organizations and technology providers will need to review the statutory text and any implementation guidance before those dates. Further details on New York’s chatbot law, including its application and enforcement, are not provided in the source material.
Key Questions
Is Microsoft’s school AI agreement a law?
No. It is a contract agreement between Microsoft and the National Academy for AI Instruction. A district can use its terms in its own Microsoft contract; the source does not describe it as a law that automatically applies to every school.
What does the agreement say about student data?
For covered education products, it bars using student, educator and administrator data to train or improve AI models. The text includes prompts, answers, uploaded files and metadata, with a narrow exception for specified safety processing that cannot be reused for general model improvement.
Does the agreement cover every AI tool used in schools?
No, not according to the supplied account. It covers products designed and marketed for education and expressly excludes general-purpose productivity and workplace tools. District contract terms determine whether the agreement applies to a particular service.
When do California’s AI laws take effect?
The source says most of the four laws signed September 30 take effect on January 1, 2027. SB 947, concerning algorithm-only discipline or dismissal decisions, is scheduled to take effect on July 1, 2027.
What remains uncertain about these safeguards?
The source does not resolve whether Microsoft’s terms apply automatically or only after a district adopts them. It also does not provide complete implementation details for the California and New York laws, or establish how regulators and courts will interpret their provisions.
Source: ThorstenMeyerAI.com
