🔍 Read the full analysis: Microsoft’s School AI Pact And California’s New Laws: What Changes For Oversight? 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 from November 1, 2026, U.S. school districts can use a contract setting limits on education-focused AI, including restrictions on data training, tracking and decisions made without human review. The pact is contractual, not a nationwide law, and California and New York have separately adopted laws addressing AI in workplaces, health care and services used by children.
Microsoft says U.S. school districts can begin using a new AI safety and privacy contract on November 1, setting restrictions on how education-focused AI handles student data, tracks behavior and makes decisions. The agreement is a contract framework, not a federal law; California and New York have adopted separate laws covering specified AI uses in workplaces, health care and services for children.
The 31-page Memorandum of Agreement was announced September 9 by Microsoft and the American Federation of Teachers (AFT) and its affiliate, the United Federation of Teachers (UFT). The source says the agreement is between Microsoft and the National Academy for AI Instruction, an educator-led training body announced by AFT in 2025, and includes spaces for other providers to sign. Microsoft says districts can incorporate its provisions into their own contracts.
For covered products, the agreement restricts use of student and educator data for model training or improvement, bans specified forms of tracking and emotional inference, and bars companion-style features designed to foster attachment or dependency. It also says covered systems cannot make decisions about students or teachers without meaningful human review or prior approval. The agreement describes a narrow exception for processing data to detect risks such as self-harm, abuse material, threats or security attacks; it says such data cannot then be used for general model improvement.
The scope is limited: the text covers AI products designed and marketed for education, while general-purpose productivity and workplace tools are expressly excluded. It also permits certain tracking exceptions with a school customer’s explicit written approval. The document sets out breach-reporting and data-deletion deadlines, but its practical effect depends on districts adopting the terms in their contracts.
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 developments put human oversight into procurement and compliance decisions, but through different mechanisms. A district that adopts Microsoft’s terms could have contractual remedies if a provider breaches them; the source says a district may end the contract and seek damages. That is distinct from a rule that automatically applies to every school or AI product.
California’s laws address specific workplace and medical uses, while New York’s law concerns certain chatbot features for children and unverified users. Together, they point to growing scrutiny of AI systems that infer personal states, affect consequential decisions or encourage emotional reliance. They do not amount to a general ban on AI. For educators, workers, parents and businesses, the immediate issue is identifying which tools and decisions fall within each measure’s scope.
Popular right now
Ninja Foodi 8-Quart DualZone 6-in-1 Air Fryer, Grey (DZ201)
Shark Navigator Lift-Away Upright Vacuum, NV360
Hisense 85 U7 Mini-LED ULED 4K UHD Smart Google TV (85U7SG, 2026)
As an affiliate, we earn on qualifying purchases.
Three Different Oversight Routes
California Governor Gavin Newsom signed four measures on September 30, 2026, according to the supplied source. AB 1883 restricts AI that reads workers’ emotions or collects neural data. SB 947 bars employment discipline or firing decisions made by an algorithm alone. AB 1979 limits AI from performing health-care functions reserved by law for licensed professionals. SB 951 requires layoff notices to disclose when AI replaced jobs. The source says most take effect January 1, 2027, while SB 947 takes effect July 1, 2027.
New York Governor Kathy Hochul signed a law on October 28, 2026, targeting unsafe companion features in chatbots used by minors and unverified users. The source describes examples including systems that pretend to be human or foster emotional dependency. These state statutes are separate from Microsoft’s voluntary-to-adopt contract framework: a school’s obligations under the pact depend on its agreement, whereas statutory duties depend on the law’s terms and applicability.
The source also reports that the AFT’s Randi Weingarten argued that enforceable provisions matter, while Microsoft President Brad Smith described the standard as setting a high bar for child privacy and AI safety. Those are supporters’ characterizations; they do not independently establish how broadly districts will adopt the terms or how the provisions will be enforced.
Adoption and Scope Still Matter
It is not yet clear how many districts will adopt the contract, which providers beyond Microsoft may sign, or how consistently its terms will be included in procurement agreements. The source describes the availability in different ways and notes uncertainty over whether protections apply automatically or only when a district requests them. The document’s wording indicates that districts can write the terms into their own contracts, but the supplied material does not resolve every implementation question.
The agreement’s education-only scope also leaves general-purpose tools outside its stated coverage, even when students or teachers use them. The source does not provide independent legal analysis of how the provisions interact with existing laws, nor does it document enforcement cases. For the state measures, the precise application to particular employers, clinicians, products or chatbot services will depend on statutory language and implementation; the source does not provide details about enforcement procedures.
District Contracts and 2027 Dates
School districts can decide whether to incorporate the agreement when negotiating Microsoft contracts, and may ask other AI providers to accept comparable terms. District procurement teams will need to identify which products are designed and marketed for education and check whether any permitted exceptions or agent features are enabled. The source says breaches must be reported within 72 hours and deleted data removed from active systems within 180 days under the agreement.
California’s stated effective dates are the next statutory milestones: January 1, 2027, for AB 1883, AB 1979 and SB 951, and July 1, 2027, for SB 947. The New York law was signed October 28, but the supplied source does not state its effective date. Further detail on implementation, district adoption and enforcement remains pending.
Key Questions
Does Microsoft’s agreement apply to every school automatically?
No automatic nationwide coverage is established. The source says districts can include the terms in their Microsoft contracts; it does not show that every district has adopted them.
What does the school AI agreement restrict?
For covered education products, it restricts using student and educator data to improve models, specified tracking and emotional inference, companion-style features, and decisions made without meaningful human review. The text includes a narrow safety-processing exception and allows certain tracking exceptions with a school’s written approval.
Does the agreement cover all AI tools used in schools?
No. The source says it covers products designed and marketed for education. General-purpose productivity and workplace tools are expressly excluded.
What do California’s new laws cover?
The four laws address worker emotion and neural-data systems, algorithm-only employment discipline or firing, health-care functions reserved for licensed professionals, and disclosure when AI replaced jobs in a layoff. The source says most take effect January 1, 2027, with SB 947 effective July 1, 2027.
When does New York’s chatbot law take effect?
The source says Governor Hochul signed the law on October 28, 2026, and describes restrictions on unsafe companion features for minors and unverified users. It does not provide the law’s effective date.
Source: ThorstenMeyerAI.com
