🔍 Read the full analysis: Do Microsoft’s School AI Pact And California’s New Laws Keep People In Control? 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 a school AI agreement will be available to U.S. districts from Nov. 1, 2026, setting contractual limits on student data use, tracking, companion features and decisions made without human review. California laws signed in September restrict certain AI uses in workplaces and health care, with most provisions taking effect in 2027. The measures do not create one broad AI ban, and their reach, enforcement and implementation differ.
Microsoft says it is making a school AI agreement available to every U.S. school district starting Nov. 1, 2026, while California has signed laws restricting several uses of AI in employment and health care. The measures draw boundaries around student data, monitoring and automated decisions, but they are not a single, uniform legal regime: the school agreement works through contracts, and California’s laws apply to particular practices on different timelines.
The agreement, called the National AI Safety & Privacy Standard, was announced Sept. 9 by the American Federation of Teachers and United Federation of Teachers with Microsoft. The source describes it as a 31-page memorandum between Microsoft and the National Academy for AI Instruction, an educator-led body announced by the AFT in 2025. It includes unfilled signature spaces for other AI providers. Microsoft says districts can use the terms in contracts for covered education products.
The terms bar providers from using student and educator data to train or improve AI models, including after de-identification. There is a limited exception for processing data to detect specified safety risks, such as self-harm, grooming, threats and security attacks; that information cannot then be used for general model improvement. The agreement also prohibits specified forms of tracking, including continuous passive attention monitoring and emotional inference, and companion features designed to foster emotional attachment or dependency.
For decisions affecting students or teachers, the agreement says AI systems must allow meaningful human review or prior approval. It also calls for student-facing agent features that act outside a provider’s systems to be off by default unless an administrator enables them. The source says breaches must be reported within 72 hours and deleted data removed from active systems within 180 days. California’s AB 1883 restricts workplace emotion-reading AI and collection of neural data; SB 947 bars discipline or firing decisions made by an algorithm alone; AB 1979 limits AI from performing tasks reserved for licensed health professionals; and SB 951 requires layoff notices to disclose when AI replaced jobs.
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 matter because they focus on consequential interactions rather than trying to stop AI use altogether. In schools, the contract could give districts a way to set conditions for education-focused products, including limits on data reuse and automated decisions. In California, statutory restrictions target particular workplace and clinical practices. For families, workers and educators, the practical question is whether a person can challenge or review an AI-influenced outcome—and whether sensitive personal information is being collected or repurposed.
But the mechanisms are different. A district would need to incorporate the agreement into a contract for it to operate as a contractual safeguard; it is not itself a law applying automatically to every school or AI product. California’s requirements, by contrast, are laws, but only cover the conduct and settings defined in each statute. The source’s broad theme—that people remain involved—is an interpretation of these separate measures, not proof that all AI decisions now require human approval.
Businesses and schools may need to identify where their tools use mood, sentiment, attention or behavioral signals, and where software contributes to discipline, termination or clinical decisions. The agreement and laws also raise procurement and compliance questions: what counts as a covered education product, what qualifies as meaningful review, and how exceptions or violations will be handled.
Products Worth Considering
DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)
Logitech MX Keys S Wireless Keyboard, Graphite
Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation
As an affiliate, we earn on qualifying purchases.
Different Rules, Different Reach
The school pact is a contractual standard, not legislation. The source says a district can incorporate its terms into a Microsoft contract and could end the contract or seek damages for a breach. Its restrictions apply to AI products designed and marketed for education, while general-purpose productivity and workplace tools are expressly excluded. The source also reports differing descriptions of whether protections apply automatically or only when a district requests them; that distinction affects how widely they operate.
California Gov. Gavin Newsom signed the four cited bills on Sept. 30, 2026. Most are scheduled to take effect Jan. 1, 2027, while SB 947 is scheduled for July 1, 2027. Separately, New York Gov. Kathy Hochul signed a law on Oct. 28 concerning chatbot companion features for minors and unverified users. The source characterizes it as restricting unsafe features, including chatbots that present themselves as human or foster emotional dependency. These developments share a focus on specified risks, but they are separate rules with different jurisdictions and coverage.
Coverage and Enforcement Questions
How the agreement takes effect across districts remains unclear from the source material. It says Microsoft is making the terms available to districts, but also notes conflicting descriptions of whether the protections apply by default or only after a district requests or incorporates them. The agreement does not cover every Microsoft tool: general-purpose productivity and workplace products are expressly outside its stated education scope.
The source does not provide the full statutory text, agency guidance, implementation procedures or enforcement records for the California laws. It also does not establish how regulators will define terms such as emotion-reading, meaningful human review or work that is legally reserved for a licensed clinician. The school agreement’s listed exceptions—including written school approval for certain tracking practices and safety-related data processing—mean its restrictions are not absolute. Whether the measures will prevent harmful outcomes in practice is not yet established.
District Contracts and 2027 Start Dates
School districts can begin reviewing Microsoft’s offered terms and decide whether to incorporate them into contracts for covered education AI products. The source says the agreement includes requirements for breach reporting, deletion and data location, but does not give a schedule for individual district adoption or identify how many districts will use it.
California’s next stated milestones are the Jan. 1, 2027, effective dates for most of the cited provisions and July 1, 2027, for SB 947. Employers, health-care organizations and technology providers will need to determine whether their systems and practices fall within each law’s scope. Further guidance, legal challenges or implementation details could clarify how the rules work; the source does not report any such developments yet.
Key Questions
Is Microsoft’s school AI agreement a law?
No. The source describes it as a memorandum of agreement that districts can incorporate into contracts for covered education products. Its contractual protections are distinct from statutory requirements.
Does the agreement cover every Microsoft AI product?
No. According to the source, it applies to AI products designed and marketed for education. General-purpose productivity and workplace tools are expressly excluded.
What does the agreement say about student data?
It bars using student and educator data to train or improve AI models, including when the data has been de-identified. A limited safety-related processing exception is described, with restrictions on using that data for general model improvement.
When do California’s cited laws take effect?
The source says most take effect on Jan. 1, 2027. SB 947, concerning discipline or firing decisions made by an algorithm alone, is scheduled to take effect on July 1, 2027.
Do these measures require a human to review every AI decision?
Not based on the information provided. The school agreement requires meaningful human review or prior approval for covered decisions affecting students or teachers, while California’s laws address specific practices. Their scope does not establish a universal human-review rule for all AI decisions.
Source: ThorstenMeyerAI.com
