🔍 Read the full analysis: How The School AI Pact And California’s New Laws Define Human Oversight 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 that from November 1, 2026, every U.S. school district can obtain a contract restricting how education-focused AI handles student data and makes decisions. California has separately enacted laws limiting workplace emotion-reading and automated employment decisions, with most provisions taking effect in 2027. The measures use different legal mechanisms and scopes, but all address when human review must remain involved.
Microsoft says it is making a new school AI contract available to every U.S. school district starting November 1, 2026, with terms that restrict student-data use, monitoring and decisions made without meaningful human review. The agreement is a contract model, not a federal law; it arrives alongside California laws setting separate limits on workplace and health-care AI, most of which take effect in 2027.
The 31-page Memorandum of Agreement was announced on September 9 by the American Federation of Teachers (AFT), the United Federation of Teachers (UFT) and Microsoft. It is between Microsoft and the educator-led National Academy for AI Instruction, with spaces for other providers to sign. Districts can include its terms in their Microsoft contracts. The source material says a district could then end a contract and seek damages for a breach, though the precise remedies would depend on the agreement and applicable contract law.
The standard bars covered education products from using student, teacher or administrator data to train or improve AI models. It also prohibits specified tracking, including continuous passive attention monitoring and emotional inference, and rules out companion-style features designed to foster emotional attachment. Covered systems may not make decisions about students or teachers without meaningful human review or prior approval. A narrow exception allows data processing to detect risks such as self-harm, abuse material, threats or security attacks; the agreement says that information cannot then be used for general model improvement.
The scope is limited: the material describes the standard as applying to AI products designed and marketed for education, while excluding general-purpose productivity and workplace tools. It also says some listed monitoring restrictions may be subject to a school’s explicit written approval. Separately, California Governor Gavin Newsom signed four AI-related bills on September 30: AB 1883 restricts workplace emotion-reading and neural-data collection; SB 947 bars employment discipline or firing decided by an algorithm alone; AB 1979 limits AI from performing work reserved for licensed health professionals; and SB 951 requires layoff notices to disclose when AI replaced jobs. The source gives January 1, 2027, as the effective date for most of the laws and July 1, 2027, for SB 947.
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 Across High-Stakes Decisions
Together, the measures put human oversight at the center of decisions affecting children, workers and patients. In schools, the contract would give districts a way to write limits into a vendor relationship rather than rely only on voluntary assurances. In California, statutory restrictions apply through state law, although the covered conduct and effective dates differ by bill.
For families and educators, the school terms speak to how student information is handled and whether software can influence placement, discipline or other consequential decisions without a person checking. For workers, the California laws target emotion monitoring and automated employment actions. For health-care patients, the relevant law draws a boundary around tasks reserved for licensed professionals. These are not blanket bans on AI: they identify particular uses and require a human role in some decisions.
The distinction matters for organizations choosing or already using software. A contract standard only has the force of the agreements districts adopt, and its reach depends on which products and terms are covered. State laws operate differently and have their own definitions, exceptions and enforcement processes. Readers should not treat the three developments as one uniform national rule.
Products Worth Considering
Apple MacBook Pro 14 2021 with M1 Max, 64GB RAM, 2TB SSD, Silver
LG 24G414B-B 24-Inch UltraGear Full HD IPS Gaming Monitor
Logitech MX Vertical Wireless Ergonomic Mouse – Graphite
As an affiliate, we earn on qualifying purchases.
Three Different Routes to Oversight
The school standard was announced by AFT, UFT and Microsoft on September 9, 2026. Its mechanism is procurement: a district may put the agreement’s requirements into a contract with Microsoft. The source says Microsoft makes the protections available from November 1, but it also notes conflicting descriptions of whether the terms apply automatically or only when a district requests or adopts them. That distinction is material for schools evaluating what protection they actually receive.
California’s measures were signed on September 30 and set legal requirements for specific workplace and health-care practices. Most are scheduled to take effect on January 1, 2027; SB 947 is scheduled for July 1. A separate New York law, signed by Governor Kathy Hochul on October 28 according to the source, addresses unsafe chatbot companion features for minors and unverified users. Its inclusion points to a wider policy focus on AI interactions with children, but it does not make the school contract or California laws broader than their stated scope.
The common thread is narrower than a general promise that AI will remain under human control. The provisions focus on identifiable practices: using sensitive data to improve models, inferring emotion or attention, letting algorithms make specified high-stakes decisions, and designing systems to foster emotional dependence.
““Anything less than legally enforceable provisions is simply a wish list.””
— AFT President Randi Weingarten
Contract Scope and Enforcement Details
Several points are not settled by the source material. It describes Microsoft’s protections as available to all districts from November 1, but says accounts differ on whether the terms attach automatically or require a district to request or incorporate them. Districts will need to examine their own contracts to determine which provisions apply.
The agreement’s limits also depend on what counts as an education-designed product, how a school exercises the written-approval exception for certain monitoring, and how breaches and damages would be handled in practice. The source states that violations must be reported within 72 hours and deleted data removed from active systems within 180 days, but it does not provide examples of enforcement or explain how disputes would be resolved.
For California, the source lists bill subjects and effective dates but does not detail every definition, exception, regulator or remedy. The laws have not all taken effect as of November 1, 2026. Their practical effects will depend on implementation and interpretation. No evidence in the supplied material establishes how many districts will adopt the contract or how providers beyond Microsoft will respond.
District Adoption and 2027 Effective Dates
School districts can review the Microsoft agreement and decide whether to incorporate its terms into their contracts. Procurement teams and school administrators will need to identify which AI products are covered, check data-handling and monitoring settings, and determine how human review will work for consequential decisions. The agreement also calls for educator training and sets operational requirements, including breach reporting within 72 hours and deletion from active systems within 180 days.
California employers, health-care organizations and technology providers will have to prepare for the relevant state provisions before their scheduled effective dates. The next dates identified in the source are January 1, 2027, for most of the laws and July 1, 2027, for SB 947. Further guidance, contract adoption and legal interpretation will show how the requirements are applied in practice.
Key Questions
Is the School AI Pact a law?
No. The source describes it as a 31-page agreement between the National Academy for AI Instruction and Microsoft. Districts can incorporate its terms into contracts; it is not a federal statute.
What does the school agreement require about human review?
Covered AI systems may not make decisions about students or teachers without meaningful human review or prior approval. The standard also says AI should support, not replace, human judgment.
Does the school standard cover every Microsoft AI product?
Not according to the supplied material. It covers products designed and marketed for education and expressly excludes general-purpose productivity and workplace tools. Districts will need to check which products and contract terms apply to them.
When do California’s AI laws take effect?
The source says most take effect January 1, 2027, while SB 947, which addresses algorithm-only employment discipline or firing, takes effect July 1, 2027.
Do these measures ban AI in schools or workplaces?
No blanket ban is described. They restrict specified uses, including certain data practices, emotion monitoring and decisions made without human involvement. The rules differ by jurisdiction, product and activity.
Source: ThorstenMeyerAI.com
