🔍 Read the full analysis: The Human Oversight Principles In Microsoft’s School AI Pact And California’s Laws 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, starting November 1, 2026, U.S. school districts can incorporate a 31-page AI agreement into contracts for education-focused products. It restricts training on student data, surveillance and decisions without meaningful human review, but it is a contract framework, not a law, and its scope and application depend on how districts adopt it.
Microsoft says it is making a new 31-page AI contract standard available to every U.S. school district starting November 1, setting limits on how covered education products can use student data and requiring meaningful human review for decisions about students and educators. The agreement is not a law: districts must incorporate its terms into contracts for the protections to apply.
The agreement, called the National AI Safety & Privacy Standard, was announced September 9 by the American Federation of Teachers and United Federation of Teachers with Microsoft. The source describes it as a memorandum of agreement between Microsoft and the National Academy for AI Instruction, an educator-led training body announced by the AFT in 2025. It includes blank signature slots for other AI providers. Microsoft says districts can put its terms into their own contracts; those contracts can provide remedies if a provider breaches the terms.
For products within its scope, the text bars providers from using prompts, answers, files, behavioral signals or metadata from students and school staff to train or improve AI models. It also prohibits specified forms of surveillance, including precise location and behavioral tracking, keystroke logging, continuous passive attention monitoring and biometric or emotional inference. The document allows exceptions to some tracking restrictions with a school customer’s explicit written approval.
The standard also bars companion-style features designed to foster emotional attachment or dependency, and says covered systems cannot make decisions about students or teachers without meaningful human review or prior approval. AI agents that act outside a provider’s systems, such as by sending messages or making purchases, must be off for students by default unless an administrator enables them. The agreement includes a narrow safety exception for processing data to detect risks such as self-harm, abuse material, grooming, threats, bullying or security attacks; it says such data cannot then be used for general model improvement.
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
Human Review at School
The standard addresses a central concern for schools adopting AI: whether software will merely assist educators or make consequential choices about children and staff. Its human-review clause is intended to keep a person involved in decisions, while restrictions on data use and monitoring set boundaries around information collected in classrooms. Those protections matter because student prompts, uploaded work and behavioral records can reveal sensitive information.
The practical force of the agreement depends on district procurement. Microsoft’s announcement makes the terms available, but the source says districts can incorporate them into their own contracts. That means schools will need to determine which products are covered, negotiate or adopt the terms, and assess whether providers comply. The document is not a blanket ban on school AI, and it does not itself establish a nationwide legal requirement.
For teachers and families, the wording also draws a distinction between educational assistance and systems that monitor attention, infer emotion or imitate a personal relationship. For technology providers, the agreement could become a set of terms school buyers ask about. Its actual reach will depend on adoption and on how its boundaries are applied to products in use.
Products Worth Considering
LG UltraGear 34GX90SA-W 34-inch WQHD Curved OLED Gaming Monitor
DJI Osmo Mobile 7P Phone Gimbal with Mic Mini Transmitter, Infinity Black
Samsung 49-Inch Odyssey G93SC QD-OLED Curved Gaming Monitor
As an affiliate, we earn on qualifying purchases.
A Contract Alongside New Laws
The school agreement is one part of a wider set of developments, but the instruments differ. The source says California Governor Gavin Newsom signed four AI-related measures on September 30, 2026: AB 1883 addresses workplace emotion-reading and neural data; SB 947 restricts employment discipline or firing decided by an algorithm alone; AB 1979 limits AI performing work reserved 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; SB 947 is scheduled for July 1, 2027.
Unlike Microsoft’s school agreement, those measures are described as state laws. Their provisions concern workplaces, health care and layoff notices rather than school purchasing contracts. The source also says New York Governor Kathy Hochul signed a law on October 28 addressing chatbot companion features for minors and unverified users. The stated restrictions concern chatbots that pretend to be human or encourage emotional dependency.
Together, these developments reflect a recurring policy question: where people should retain authority when AI systems influence education, employment, health care or children’s interactions. That shared theme does not make the rules interchangeable. Each has a different legal basis, affected group and scope, and the school standard applies through contracting rather than automatically as a law.
Scope and Adoption Questions
The available source describes the standard’s rollout in two ways: Microsoft is making it available to every district, while the agreement is described as terms districts can write into their contracts. It is therefore not clear that protections apply automatically to every district or product on November 1. The source also says general-purpose productivity and workplace tools are expressly excluded, limiting the standard to AI products designed and marketed for education.
Other implementation details remain unresolved in the material provided, including how districts will verify compliance, how many districts will adopt the terms, and which additional providers will sign the agreement. The text allows written school approval for exceptions to some tracking restrictions, but the source does not explain how often such approvals might be sought or granted. The safety exception is narrowly described, though its oversight in practice is not detailed.
The California and New York measures are summarized in the source, but their full statutory language and implementation guidance are not supplied here. Their practical effects, enforcement processes and any future legal challenges are not established by the available information.
District Contracting and Compliance
Starting November 1, districts can consider incorporating the standard into contracts for covered education AI products. Their next steps may include reviewing existing agreements, identifying which tools fall within the education-product definition, and deciding whether to adopt the data, monitoring and human-review terms. The document also sets out operational requirements, including breach reporting within 72 hours and deletion of data from active systems within 180 days.
Schools and providers will need to clarify how the terms work in practice, especially where a district seeks a written exception or a product combines educational functions with general-purpose tools. The agreement’s blank signature spaces leave open whether other providers will sign or offer comparable commitments.
For the state measures, the dates identified in the source are the next clear milestones: most of California’s provisions are scheduled for January 1, 2027, with SB 947 scheduled for July 1. The source does not provide an implementation timetable for New York’s law. Further guidance, district contract decisions and provider participation should show how much of the stated human-oversight approach reaches schools and users.
Key Questions
Is Microsoft’s school AI standard a law?
No. It is a contract standard, not legislation. Districts must incorporate its terms into contracts for the protections to apply as described in the source.
What does the agreement prohibit?
For covered education AI products, it bars training or improving models with student and school-staff data, restricts specified tracking and emotional inference, prohibits companion-style features designed to foster dependency, and requires meaningful human review for decisions about students or educators.
Does the standard cover every Microsoft AI product used by a school?
The source says it covers AI products designed and marketed for education. General-purpose productivity and workplace tools are expressly excluded, so the scope of any particular product may need to be clarified.
Can student data ever be processed for safety reasons?
The agreement describes a narrow exception for detecting risks such as self-harm, abuse material, grooming, threats, bullying or security attacks. It says that data cannot then be used for general model improvement.
When do the California laws take effect?
According to the source, most of the four measures take effect on January 1, 2027. SB 947 is scheduled to take effect on July 1, 2027.
Source: ThorstenMeyerAI.com
