🔍 Read the full analysis: The Rules Behind Human Control In Microsoft’s School AI Pact And California’s New Laws on ThorstenMeyerAI.com
Get hardware and tech essentials 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 a 31-page AI safety agreement is available to U.S. school districts starting Nov. 1, 2026. The contract restricts how covered education AI can use student data and requires meaningful human review for decisions, but it is not a law and its scope does not include general-purpose tools.
Microsoft says that starting Nov. 1, 2026, every U.S. school district can access a contract that restricts how covered AI products handle student information and requires human review before decisions about students or educators. The 31-page agreement is a contract framework, not a new law, and districts must incorporate its terms into their own agreements for the protections to be enforceable.
The agreement, called the National AI Safety & Privacy Standard, was announced Sept. 9 by Microsoft and the teachers’ unions AFT and UFT. Source material describes it as a memorandum between Microsoft and the National Academy for AI Instruction, an educator-led training body announced by AFT in 2025, with signature spaces for other AI providers. Microsoft says districts can use its terms in contracts for education-focused AI products.
Under the text, providers may not use prompts, responses, uploaded files, behavioral signals or metadata from students and school staff to train or improve AI models. The restriction is described as permanent, including after a contract ends. A narrow exception permits processing for specified safety concerns, including threats of violence, self-harm risk, grooming, bullying and security attacks; that data may not then be used for general model improvement.
The agreement also bars specified forms of student tracking and emotional inference, including continuous passive attention monitoring, long-term profiling and biometric collection. The source says the list allows exceptions with a school customer’s explicit written approval. It prohibits companion-style features designed to encourage emotional attachment or dependency and says covered systems cannot make decisions about students or teachers without meaningful human review or prior approval. Some agent-like features that act outside the provider’s systems must be off for students unless an administrator enables 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
Contract Terms Put Districts in Control
The agreement’s reach depends on school procurement decisions, rather than a nationwide legal mandate. If a district puts the terms into its Microsoft contract, the source says it can terminate the agreement and seek damages for a breach. That makes contract language a practical lever for districts seeking limits on data use and automated decisions, while leaving each district responsible for deciding whether and how to adopt it.
For educators and families, the main change is a stated boundary between AI assistance and decisions that affect people. The standard says educators should retain meaningful oversight, and the Academy will instruct educators not to use AI alone for automated grading, discipline, placement or emotional assessment. These provisions do not prohibit schools from using AI; they define restrictions for covered products and uses.
The development also sits alongside separate state actions. California has enacted laws addressing workplace monitoring, algorithmic discipline and health-care AI, while New York has adopted rules on chatbots for children. The measures share concerns about human decision-making and sensitive data, but their legal status, covered settings and enforcement mechanisms differ. They should not be treated as one unified national rule.
Products Worth Considering
Logitech MX Brio Ultra HD 4K Webcam, Graphite
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
As an affiliate, we earn on qualifying purchases.
How the School Standard Took Shape
AFT and UFT announced the standard with Microsoft on Sept. 9, 2026. Microsoft’s stated Nov. 1 availability date is the next step: the company says districts across the country can use the agreement. The source describes the document as a memorandum with empty signature slots for other AI providers, but the detailed terms supplied focus on Microsoft and education AI.
Separately, California Gov. Gavin Newsom signed four bills on Sept. 30. As described in the source, AB 1883 restricts AI emotion reading and neural-data collection at work; SB 947 limits discipline or firing decided by an algorithm alone; AB 1979 restricts AI from performing health-care functions reserved for licensed professionals; and SB 951 requires layoff notices to disclose when AI replaced jobs. Most are set to take effect Jan. 1, 2027, while SB 947 is scheduled for July 1, 2027.
New York Gov. Kathy Hochul signed a separate law on Oct. 28 concerning chatbots used by minors and unverified users. The source characterizes it as restricting unsafe companion features, including systems that pretend to be human or encourage emotional dependency. The three developments are related in theme, but the school agreement is contractual, while the California and New York measures are laws.
Adoption and Coverage Remain Open
It is not clear from the source whether Microsoft’s offer automatically places the terms in every district’s contract or requires each district to request and adopt them. The source itself says Microsoft’s description, the agreement and press reports differ on this point. The practical answer will matter: access to a standard contract is not the same as automatic coverage.
The agreement is limited to AI products designed and marketed for education. General-purpose productivity and workplace tools are expressly excluded, so the source does not establish that the restrictions apply whenever a student or teacher uses a broadly available AI tool. It also describes some tracking exceptions as possible with written school approval; the scope and use of that provision may affect how the restrictions work in practice.
The material does not provide a full account of enforcement procedures, any district’s adoption of the terms, or whether other AI companies will sign on. It also does not establish how the California and New York laws will be applied in particular cases. Those details remain unsettled based on the information provided.
District Contracts and Effective Dates
Districts can now consider whether to place the Microsoft standard’s terms in their agreements. The next practical indicator will be whether districts adopt the language and how Microsoft and schools document compliance, including the stated 72-hour breach-reporting deadline and requirement to delete data from active systems within 180 days.
California’s listed laws are scheduled to begin taking effect on Jan. 1, 2027, with SB 947 following on July 1, 2027. New York’s chatbot law will also require follow-up on its implementation and enforcement. Until district contracts and state rules are applied, the extent of real-world protection—and the consequences for violations—remains to be seen.
Key Questions
Is Microsoft’s school AI standard a law?
No. It is a contract framework, not legislation. According to the source, districts can make its provisions enforceable by incorporating them into their own contracts.
What does the agreement say about student data?
It bars covered providers from using student and educator prompts, files, responses and related data to train or improve AI models. A limited exception covers specified safety processing, which the source says cannot be redirected to general model improvement.
Does the standard cover every AI tool used at school?
The source says it covers AI products designed and marketed for education. General-purpose productivity and workplace tools are excluded, and it is not clear whether districts receive the terms automatically or must adopt them.
What does human review mean under the agreement?
The text says covered systems cannot make decisions about students or teachers without meaningful human review or prior approval. The Academy will also instruct educators not to rely on AI alone for grading, discipline, placement or emotional assessment.
When do the California laws take effect?
According to the source, most of the four California laws signed Sept. 30 are scheduled to take effect on Jan. 1, 2027. SB 947, concerning algorithmic discipline or firing, is scheduled for July 1, 2027.
Source: ThorstenMeyerAI.com
