🔍 Read the full analysis: Who Stays In Charge When Schools Use AI? What The Pact And Laws Say on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Microsoft says that from November 1, 2026, its school AI protections are available to U.S. school districts through contract terms, not a new federal law. The terms restrict data use, monitoring, companion features and decisions made without human review; separate California and New York laws address workplace and child-facing AI.
Microsoft says U.S. school districts can use new AI contract protections starting November 1, including limits on using student data to train models and requirements for human review of decisions. The terms are part of an agreement with an educator-led organization, not a new law; their reach depends on how districts adopt them and which products they cover.
The 31-page Memorandum of Agreement was announced September 9 by the American Federation of Teachers (AFT), its United Federation of Teachers affiliate, the National Academy for AI Instruction and Microsoft. The source describes the Academy as an educator-led training body announced by AFT in 2025. Microsoft says the protections are available to every U.S. school district from November 1. The document also has blank signature spaces for other AI providers.
For products designed and marketed for education, the agreement bars providers from using prompts, answers, uploaded files, behavioral signals and metadata from students and educators to train or improve AI models. It also prohibits specified forms of tracking, including continuous passive attention monitoring, and companion-style features intended to foster emotional attachment. The document allows a narrow data-processing exception for safety concerns such as self-harm risk, abuse material, grooming, threats, bullying or security attacks; it says that data cannot then be used for general model improvement.
The agreement says covered systems may not make decisions about students or teachers without meaningful human review or prior approval. It also calls for student-facing agent features to be off by default, breach reports within 72 hours and deletion from active systems within 180 days. Some monitoring restrictions allow an exception with explicit written approval from the school customer. The terms can be incorporated into district contracts, according to the source; a district could then use contractual remedies if a provider breaches 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
How Contract Terms Keep People Involved
The agreement puts practical limits on who controls school AI: providers cannot treat student information as general training material, and software is not meant to make consequential decisions without people reviewing them. For families, educators and administrators, those terms address concerns about privacy, profiling and automated judgments in classrooms.
But a contract is not a universal rule for every AI tool. Its protections apply to covered education products and depend on districts adopting the terms. The source says general-purpose productivity and workplace tools are expressly excluded. That distinction matters for schools buying bundled services: a tool used by school staff may not fall under the same provisions simply because it is used in an educational setting.
The broader pattern is a human-review boundary, but the mechanisms differ. A contract can bind a provider and customer that accept its terms; statutes apply according to their scope and effective dates. Schools will need to understand which protections are written into their own agreements rather than assume every classroom AI product follows this standard.
Products Worth Considering
DJI Osmo Mobile 8 Gimbal with Mic Mini Transmitter, Infinity Black
Amazon Kindle Scribe (64GB) with Premium Pen, Metallic Jade
Bose SoundLink Plus Portable Bluetooth Speaker, Black
As an affiliate, we earn on qualifying purchases.
Three Different Routes to Human Oversight
The school agreement is part of a wider set of developments described in the source. On September 30, California Governor Gavin Newsom signed four bills addressing AI in employment and health care. The source says AB 1883 restricts AI emotion-reading and collection of neural data at work; 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 say when AI replaced jobs.
Most of those California measures are scheduled to take effect January 1, 2027, while SB 947 is scheduled for July 1, 2027, according to the source. Separately, New York Governor Kathy Hochul signed a law on October 28 covering chatbots used by minors and unverified users. The source describes it as prohibiting specified “unsafe” companion features, including chatbots that pretend to be human or foster emotional dependency.
These developments do not create one national AI rule. They use different tools—school procurement contracts and state laws—and cover different settings. The common thread is limiting AI’s role in reading personal signals, shaping relationships with users or making consequential decisions without human involvement.
“AI tools support educators and students. They do not replace human judgment.”
— The school agreement
Which Schools and Tools Are Covered
How the terms become operative for each district is not fully clear from the source. It says districts can write the protections into their Microsoft contracts, but also reports differing descriptions of whether the offer applies automatically or only when a district asks. The agreement’s practical reach will depend on the contract language a district adopts and the products it covers.
The source identifies general-purpose productivity and workplace tools as excluded, but does not provide a full product-by-product list. It also does not explain how districts will handle the agreement’s written-approval exception for certain tracking restrictions, or how compliance will be checked in practice. The laws’ eventual implementation and any interpretation by regulators or courts are not addressed in the material.
District Contracts and State Effective Dates
From November 1, 2026, districts can seek to use the Microsoft agreement’s provisions in their contracts. The next practical step is for school systems to identify which AI products are covered, review existing terms and decide whether to adopt the protections. The source says the agreement sets breach-reporting and deletion timelines, but it does not identify a schedule for district adoption or provide evidence of compliance outcomes.
California’s measures are scheduled to take effect on the dates specified in each law—mostly January 1, 2027, with SB 947 on July 1, 2027. New York’s chatbot law was signed October 28; the source does not specify its effective date. For all three areas, the key developments to watch are the final contract terms districts use, guidance on the laws, and how schools, employers and providers apply the rules to particular AI systems.
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. Districts can incorporate its terms into their own contracts; it does not automatically create a rule for every AI provider or school.
What student data can covered AI products use for training?
The agreement bars using covered student and educator prompts, responses, files, behavioral signals and metadata to train or improve models. It sets out a limited safety-processing exception, while prohibiting that information from flowing into general model improvement.
Does the agreement cover every AI tool used in a school?
No. The source says it covers AI products designed and marketed for education and excludes general-purpose productivity and workplace tools. A district’s contract and the product’s scope matter.
When do the California AI laws take effect?
According to the source, most take effect January 1, 2027. SB 947, concerning discipline or firing decided by algorithms alone, is scheduled to take effect July 1, 2027.
What remains uncertain about implementation?
The source reports differing descriptions of whether Microsoft’s offer applies automatically or requires a district request. It also does not give a complete list of covered products or explain how compliance will be monitored across districts.
Source: ThorstenMeyerAI.com
