K-12Latest editionIssue 65
K–12
A Nov. 10 Stanford SCALE update of 52 causal studies finds K-12 generative AI helps more when students use it to learn than to produce work, but neither use is consistently beneficial.
Full research summary
A Nov. 10 Stanford SCALE update of 52 causal studies finds K-12 generative AI helps more when students use it to learn than to produce work, but neither use is consistently beneficial. It cites a large NBER trial of more than 6,000 Tennessee middle-school students in which an AI math tutor improved post-mistake correctness but slowed practice and showed only modest, marginal delayed-test gains. A Nov. 10 Texas 2036 report adds that AI use outpaces district policy and that districts, not vendors, carry student-privacy duties. Evidence is mostly preprints and non-U.S. or mixed-age.
Read the edition Previous: Issue 64, November 9All K–12 editions
What this stream covers
Primary/secondary schools, districts, educators and school-age learners. Cover teaching, learning, district operations, safety, privacy and access. Distinguish assisted output from retained learning and evaluate developmental risks.
- Evidence records
- 3
- Cross-source patterns
- 1
- Also published November 10
- AnthropicAWSCampus OperationsCiscoCitizen Engagement & ServicesCrowdStrikeDellEmergency ServicesEverpureGoogleHealth & Human ServicesLocal GovernmentMicrosoftNetAppNVIDIAOpenAIPalo Alto NetworksPublic SafetyRed HatResearch UniversitiesServiceNowState GovernmentStudent SuccessTransportationUtility ManagementVAST Data
- Subscribe
- Atom feed for K–12
What the evidence supports, and what it leaves open
1 pattern this edition will stand behind, each carried by at least two independent sources. Every one arrives with the question it has not answered.
- Learning effects depend on how AI is structured, not whether students have it
Operating questionWhich specific student uses will the district permit, and what delayed unaided measure will show whether a tool is helping learning?
Research through your lens
Every resource includes source evidence and takeaways for all three roles.
Evidence in this micro-vertical
281 resources
Follow the outcomes
281 resources across outcomes in your selection. Counts include all outcomes.
Refine by evidence type and topic
- Source
- Claude Help Center
- Published
- Undated help article (retrieved November 10, 2026; feature launched November 10, 2026)
Claude Help Center: Whiteboards beta is unavailable to HIPAA-ready organizations, not covered by Claude Code zero data retention, and not yet in Claude for Teachers
Anthropic's Help Center says Claude Whiteboards, a shared board Claude organizes from spoken, typed or sketched input, is in beta, on by default for Team plans and off by default for Enterprise until enabled. At launch it is unavailable to organizations with HIPAA-ready configurations, is not covered by zero data retention in organizations that have ZDR for Claude Code, does not use connected-app context, and is not yet available in Claude for Teachers. Boards can be shared with outside invitees only when admins allow it and can be sent to Lucid, Miro or Zoho Vani.
Limitations & uncertainty
Undated vendor help article; describes launch-state limits that may change; does not separately document Claude Meetings retention. Cross-tagged health-human-services (HIPAA-ready exclusion) and k12 (not yet in Claude for Teachers).
NBER working paper: randomized trial of an AI math tutor with more than 6,000 Hamilton County, Tennessee middle-school students
Students were randomized to AI versus computer-assisted-learning-only support, mastery versus non-mastery progression and math topic. AI students attempted fewer questions and moved more slowly but were more accurate per attempt, and AI improved next-attempt correctness after mistakes. On a delayed test one week later, gains were modest: about three percentage points on the practiced Exercise 1 item for mastery students, only marginally significant, with no overall delayed-learning gain. Mastery alone did not improve delayed learning.
Limitations & uncertainty
Working paper, not peer reviewed. One short session and a one-week delay; the authors say it does not estimate effects of repeated use. The delayed-test effect is marginally significant and concentrated on one item. Platform-specific (NUMI), single-region study, and the tutor was a guard-railed design, so results may not carry over to general chatbots. Sample size is stated as more than 6,000 in the abstract.
Stanford SCALE AI Hub for Education: The Evidence Base on AI in K-12, a November 2026 review of 52 causal studies
An update to the March 2026 review adds 32 strong causal papers published October 2025 to June 2026, for 52 in total. Reviewers conclude AI is more likely to support learning when students use it to understand material than to produce work, but neither use consistently improves learning, and offloading tasks can reduce learning. AI can improve learning experience and confidence without improving outcomes, and teacher-facing AI can cut planning time or raise feedback quality but one new study found it reduced student motivation and confidence.
Limitations & uncertainty
The authors say findings should not be treated as broadly generalizable across U.S. K-12 settings; the U.S. K-12 share of studies is limited. Many sources are non-peer-reviewed preprints, the repository is arXiv-weighted so relevant journal studies may be missed, and LLM-assisted screening may have excluded relevant studies. Some included studies involve undergraduates or adults rather than school-age students. This is a secondary synthesis; effect sizes are not pooled.
Texas 2036: In Ctrl of AI in Schools, a Texas policy report on AI governance, student safety, learning evidence and AI literacy
The report finds AI is already in Texas classrooms but largely unguided: in fall 2025, 86% of school and district leaders reported teacher AI use and 61% student use, yet only 29% reported a district policy on student AI use and 20% one on teacher and staff use. It argues FERPA obligations fall on schools rather than vendors, that student PII includes indirect identifiers, and that learning evidence is genuinely mixed. It recommends a state baseline for responsible use, pre-classroom vetting of tool safety and efficacy, educator AI literacy and student AI literacy.
Limitations & uncertainty
Advocacy-oriented policy report from a Texas organization, not a primary evaluation. It states that most underlying evidence comes from outside the United States, reviews draw more on higher education than K-12, causal studies are few and some not peer reviewed, and findings are early signals. Adoption and policy percentages are self-reported leader survey figures. Legal statements are the authors' interpretation, not legal advice.
- Source
- Portland State University News
- Published
- November 9th 2026
Portland State $2M Gates/Learning Commons project to benchmark how Claude and other models give feedback on grades 6-12 data reasoning
Portland State University researchers received a two-year, $2 million Gates Foundation and Learning Commons grant to build a benchmark for how AI systems, explicitly including Claude and GPT-6 Astra, respond to grades 6-12 students' data-reasoning work. The benchmark will use 1,200 de-identified iterative samples from partner districts and score whether models interpret student reasoning, choose an appropriate instructional move and preserve the student's thinking. Researchers say it will give districts an independent way to examine capabilities before adoption. No results yet.
Limitations & uncertainty
Announcement of planned research; no results; partner districts and model versions not named; release is from the university itself. Cross-tagged k12 (benchmark built from district student work) and research (university-led evaluation infrastructure).
- Source
- The Harbinger (Shawnee Mission East High School)
- Published
- November 9, 2026
Shawnee Mission School District (Kan.) ICT team pilots a Claude system to triage help desk and website-unblock requests; one test took 14 hours
The Harbinger reports that Shawnee Mission School District's ICT team has been piloting a Claude-based system for a few weeks to speed triage of IT help desk requests, with a four-to-five-person team serving more than 26,000 students and 4,000 employees. The pilot is early; a recent test examining one potential website-unblock request took 14 hours. Leaders say it could free staff to review changing app user agreements and age limits, while a teacher criticized the related app restrictions and a student called the pilot and restrictions confusing. ICT leadership declined to comment.
Limitations & uncertainty
Student newspaper report; relies on board recaps and interviews; ICT leadership declined comment; no detail on Claude product, configuration, data handling, or what the 14-hour test involved or compared against. Cross-tagged k12: district IT operations.
School Data Leadership Association explainer: Microsoft's Nov. 5 education licensing changes, A0 limits, pooled storage maths and why Copilot Chat is not A7 Copilot
SDLA's explainer, based on Microsoft's Nov. 5, 2026 announcement and guidance, tells districts that A0 is a limited, conditional no-cost student option (Office apps, OneDrive and Minecraft Education; no Exchange, SharePoint or Teams), requiring full paid faculty coverage on A2 or higher and active Windows use, with up to 20 A0 student licences per eligible paid faculty licence. It states that A7 builds on A5 with Microsoft 365 Copilot plus identity, device and AI-agent management, and warns that Copilot Chat and Microsoft 365 Copilot are different offerings and that student access depends on eligibility and admin settings. It explains the 100 TB base storage allocation goes away at the applicable renewal, gives per-licence pooled storage contributions (e.g., A3 faculty 75 GB, A2 student 4 GB) and a worked example of 57,500 GB for 500 A3 faculty plus 5,000 A2 students, and advises reviewing AI features' access to district information and controls before enabling them. Independent interpretation: 'an unchanged faculty suite price does not guarantee an unchanged district bill.'
Limitations & uncertainty
Secondary analysis of a vendor announcement; publication date inferred from the cited Nov. 5 announcement (page shows no explicit post date, so publicationDate is null); figures depend on Microsoft guidance that may change; written by a California-based nonprofit with a K-12 focus.
- Source
- Torrance USD Board agenda (Diligent Community)
- Published
- Undated board agenda item (agreement term begins October 19, 2026)
Torrance Unified (Calif.) board item: three-year, $1.39M GuidePoint agreement for CrowdStrike Falcon MDR, identity, exposure management and SIEM via NASPO
A Torrance Unified board agenda item requests approval of a three-year agreement with GuidePoint Security for CrowdStrike software, support and services totaling $1,387,202.72 from the General Fund/Technology budget, for Oct 19, 2026 through Oct 18, 2029. The Falcon scope listed is managed detection and response, endpoint protection, identity threat protection, exposure management, security information and event management, threat monitoring and extended data retention. The purchase piggybacks NASPO ValuePoint California Contract 7-17-70-40-05 (Master AR2472) under Public Contract Code 20118 and a June 16, 2025 board authorization.
Limitations & uncertainty
Meeting date and approval outcome not shown on the opened page; no AI features named; no efficacy metrics.
German one-year study of 4,062 secondary students: AI use for finishing tasks is linked to lower later effort, and neither use type predicts test scores
The article summarizes a preprint posted Nov. 6, 2026 that followed 4,062 pupils in 136 Baden-Württemberg secondary schools through 2023/24 with standardized tests in six subjects at start and end of year. By year end 62% reported using AI for homework and 65% for exam preparation; about 5% never used it. Lower-scoring pupils went on to use AI more. Neither task-completion nor learning-oriented use predicted later test scores in any subject. A profile weighted toward task completion predicted lower later effort in four of six subjects, while learning-oriented use was linked to higher effort in mathematics, chemistry and physics. The authors and article caution that the work is a preprint and use is self-reported.
Limitations & uncertainty
Preprint, not peer reviewed; data from 2023/24 before newer tools; self-reported use; German sample; observational so no causal claim; article is secondary and the primary paper was not inspected.
Seattle Public Schools closes a Gemini AI loophole and keeps Copilot staff-only, but says it cannot turn off Word's AI rewrite features for students
The Seattle Times reported on Nov. 7, 2026 that Seattle Public Schools announced at a special School Board meeting the previous Wednesday (Nov. 4) that it will block YouTube for K-5 students and Minecraft in most classes, and close a route by which students reached Google's Gemini chatbot through search AI summaries via a web-filter vendor update. Claude and Grok were already blocked on the network. Source findings relevant to Microsoft: the district lists approved student AI tools (MagicSchool AI, Canva Magic Write, Adobe Firefly) while Microsoft Copilot is available for staff only; when a student board representative asked whether the district could turn off AI features that rewrite highlighted text in Microsoft Word, the digital learning manager said they cannot. Roughly 40–50 parents and advocates rallied for blocking all generative AI chatbots for K-12 students, including chatbots built into district-approved tools, until independent research shows safety and learning benefit; a board vice president called claimed ed-tech benefits 'very based on vibes'. Interpretation: Microsoft AI embedded in core Office apps is outside some districts' practical on/off control, which complicates staff-only Copilot policies.
Limitations & uncertainty
News reporting of meeting statements; the claim that Word AI rewrite cannot be disabled is a district staff answer, not verified against Microsoft admin documentation, and may depend on licence, app version or policy settings. Education Lab is Gates Foundation-supported (disclosed). No learning-outcome data.
Gothamist: NYC’s Google Search/AI standoff deepens after mayoral one-year ban on most school AI programs
Gothamist reports Google and NYC Public Schools are in a standoff after Google indicated schools that ban AI could lose Search. Mayor Zohran Mamdani, who in September announced a one-year ban on most AI programs and tighter screen limits for youngest students, accused Google of using monopoly power to force AI into schools. Officials cite 1.3 million Chromebooks (≈500,000 student-assigned). The city has disabled Gemini chatbot and AI Mode on student Chromebooks and asked Google to disable AI Overview that still appears; Google publicly framed Gemini/Notebook/Search as off by default with school choice and said AI Mode/Overviews will emphasize step-by-step guidance. Parent advocates warn disentangling from Google Classroom/Docs/Search is extremely hard.
Limitations & uncertainty
Google did not comment to Gothamist; Search-loss scenario may evolve with Google’s off-by-default clarification; lawsuit is allegation only.
Michigan DOE uses CCSSO Digital Excellence grant to launch multi-state AI Together on AI, screens, privacy, integrity and well-being
On Nov. 6, 2026, MDE announced $53,000 in second-round CCSSO Digital Excellence Collaborative Community funding (Ballmer Group) to begin multi-state work with Delaware, Kentucky, Nebraska, New Mexico, North Carolina and Wyoming. AI Together: Connecting States, Schools, and Families will develop engagement strategies on AI, screen time, data privacy, academic integrity and student well-being; a companion Student Voice project will surface student and educator barriers to equitable digital learning. Materials are intended for sharing with all states. A prior $40,000 mini-grant produced multilingual parent tech-support videos in GoOpen Michigan.
Limitations & uncertainty
Announcement of work beginning; deliverables, evaluation methods and timelines not specified; $53,000 budget is limited relative to statewide need.
South Portland board to vote on PK–8 generative AI moratorium through 2026-27 amid missing policy and staff training mandate
The Press Herald reports the South Portland School Board will vote Monday on a moratorium on AI use through the 2026-27 school year for pre-kindergarten through 8th grade until a policy is developed; high school is excluded. If adopted, all staff would complete two hours of AI awareness and three hours of AI literacy training by year-end, then use only district-sanctioned tools on school devices, with no requirement to use student-facing AI. The district currently has no official AI guidelines; MagicSchool is already deployed as a closed, teacher-controlled tool; exceptions would cover disabilities, English learners, and approved coding/simulations. The proposal lacks an enforcement plan.
Limitations & uncertainty
Vote pending; enforcement unspecified; article’s general claim about post-AI test-score drops lacks a named study citation.
Broward County Public Schools names first Microsoft AI/Copilot Innovation Award winners with school-level usage counts
On Nov. 5, 2026 Broward County Public Schools published that Pompano Beach High, Peters Elementary and Indian Trace Elementary are the first monthly winners of its Microsoft AI/Copilot Innovation Recognition Program, each earning $4,000 and advancing to a School of the Year competition after recognition at the Nov. 4 School Board Workshop. Source findings: fourteen schools applied in September; Pompano Beach High staff logged more than 5,300 Copilot interactions since April across ELA, science, social studies, math, electives, administration and counseling for lesson drafting, differentiation and family communications; Peters Elementary used Copilot to merge teacher schedules, MTSS data and progress-monitoring lists into an RTI meeting schedule prioritizing Tier 3 and third-grade students; Indian Trace delivered schoolwide Pre-K–5 Copilot training with a shared Goal–Context–Source–Expectations prompting method and a privacy-first approach to small-group planning. Interpretation: a US K-12 district is publishing school-level Copilot usage and operational use cases under a Microsoft-funded recognition program, without independent learning-outcome measures in this article.
Limitations & uncertainty
District communications, not an independent audit; Microsoft funds the $4,000 awards; interaction counts are self-reported; no student-learning outcome metrics; student Copilot access is outside this article's scope.
BusinessWorld: Microsoft, DepEd and DICT launch AI literacy credentials for 2,000 teachers atop Copilot access
BusinessWorld reported on Nov. 5, 2026 that Microsoft Philippines partnered with DepEd and DICT to train and accredit 2,000 public school teachers using Microsoft Elevate for Educators, the AI Classroom Toolkit and Behind the Chat. Source finding: the initiative builds on an earlier agreement to grant 1 million educators access to Microsoft Copilot. Vendor claims relayed by the paper include Microsoft internal research that about nine in ten educators have used AI for school tasks. No independent student-outcome or safety-impact evaluation is presented.
Limitations & uncertainty
Philippines legal/education context; Copilot seat count and educator-usage stats are vendor claims; no measured classroom outcomes.
Hanover Research K-12 AI survey of about 300 parents and staff finds growing student AI use alongside guidance, training and data-safeguard gaps
EdTech Magazine reports that Hanover Research surveyed nearly 300 parents and K-12 staff. More than half said students use AI in schoolwork, most often for writing help and generating media; nearly 7 in 10 staff use AI sometimes or often in their own work. Most respondents said students are told when to use AI but not always how; half of staff said they lack time to train students. Just over half thought data safeguards are adequate and about the same share trust AI output. Nearly 7 in 10 staff want more AI training and only 37% feel very comfortable showing students how to use AI for schoolwork. Hanover recommends a living districtwide AI framework, data governance language and recurring professional learning.
Limitations & uncertainty
Small sample (about 300), unclear sampling and representativeness, perceptions rather than outcomes, vendor-authored recommendations; the underlying report was not opened.
EdWeek Market Brief: SETDA survey finds AI still the top state ed-tech priority as professional development gaps widen
In a Nov. 5, 2026 Education Week Market Brief article, Scott Elliott reports that school technology leaders ranked AI as their top priority for the second straight year in SETDA’s fifth annual State EdTech Trends Report. The report is based on surveys of ed-tech directors, state chiefs, chief information officers, and other state education agency leaders in 45 states, plus spotlight reports on North Carolina, Tennessee, Illinois, and Colorado. The article states that 53% of states have published guidance on buying AI-enabled tools, up from about 40% last year. For states that offer purchasing guidance, nearly all of those surveys said theirs include safety measures, and about 8 in 10 have an evidence-based quality indicator, up from 45% the prior year. Of the survey respondents, 62% cited professional development as the top unmet technology need for educators, up 15 points from the prior year and the second-biggest unmet need overall behind only funding (72%). The reporter’s lede characterizes AI-related professional development demand as growing fast; the 62% figure in the body is specifically PD as the top unmet technology need for educators, not a separately broken-out “AI-only PD” percentage. Authors quoted in the article write that the pressure on state leaders is no longer to get technology into students’ hands but to show that technology already there helps students learn, keeps them safe, and is worth what states and districts are paying for it. A Tennessee spotlight item says the state education department and commissioner’s office have undertaken a collaborative effort to write statewide AI guidance for schools. This Nov. 5 article was not in the archive; it is included because it reports current 45-state survey measures of AI priority, purchasing-guidance adoption, and educator PD gaps that remain newly relevant to state and district AI operating decisions after the last completed run.
Limitations & uncertainty
This is a three-minute secondary article dated Nov. 5, 2026, about a SETDA survey whose own press release (inspected separately and not used as this resource) is dated October 29, 2026—before the last completed run. The Market Brief page does not state the survey’s response count, weighting, or exact fieldwork dates. Do not treat the reporter’s lede phrase “AI-related professional development” as identical to the 62% body statistic, which refers to PD as the top unmet technology need for educators. Spotlight items on Tennessee, Colorado, North Carolina, and Illinois are brief illustrations, not evaluated case studies. Figures such as “about 40%” and “about 8 in 10” are the article’s own approximations.
McGraw Hill opens a beta MCP connector so districts can use licensed K-12 curriculum inside Claude for Teachers
Investing.com reports McGraw Hill will let school-district customers access licensed K-12 curriculum inside Claude for Teachers through a Model Context Protocol connector, currently in a beta open to interested districts. Educators could draft lesson plans and personalized materials from their district's McGraw Hill content. The piece also notes TeachFX is already in the Claude for Teachers ecosystem and that McGraw Hill separately selected Claude as an internal employee productivity tool. No measured classroom outcomes, student-privacy evaluation, or district deployment results are reported.
Limitations & uncertainty
Vendor collaboration reported by a financial news site; original BusinessWire release was inaccessible (HTTP 403) during this run. No independent evaluation of instructional quality, accessibility, or privacy. Beta scope and eligibility are not fully specified here.
Seattle Public Schools bans Gemini districtwide, tightens elementary screens, and sketches a grade-banded AI literacy ladder amid parent pushback
At a Nov. 4 board meeting, Seattle Public Schools said Google Gemini is now banned for all students after a prior K-2 ban and a student loophole; YouTube will be blocked for K-5 and Minecraft for K-12 starting the next Monday, with curriculum exceptions for Minecraft. Officials described purchased MagicSchool AI and Microsoft Copilot and a developmental AI literacy vision (no classroom AI in K-2; learn-about-AI from grade 3; teacher-guided error/bias work by grade 6; high-school argument stress-testing and draft feedback). Board Director Evan Briggs said the presentation showed no data that MagicSchool raised reading scores, and Seattle Families for Intentional Tech rallied for a pause on student-facing generative AI.
Limitations & uncertainty
Single report; technical loophole unspecified; no independent learning-outcome evaluation of MagicSchool or Copilot in SPS.
Microsoft announces Microsoft 365 Education portfolio overhaul: A7 bundles its most advanced AI, Office 365 A1 and Student Use Benefit retire Jan. 1, 2028
Microsoft's licensing news page dated Nov. 5, 2026 describes an updated Microsoft 365 Education portfolio for K-12 and higher education: faculty and student editions with or without Teams, starting with a no-cost student option; Microsoft 365 Education A7 positioned as offering its most advanced AI, security, governance and device management capabilities; added faculty security and device capabilities at no additional cost; storage that scales with faculty and student populations; and Microsoft Classroom as a dedicated teaching and learning experience while Teams for Education remains. Key dates: pricing updates for volume-licensed online services begin May 1, 2027; on Jan. 1, 2028 Office 365 A1, the Student Use Benefit and free Exchange Online for Alumni retire alongside storage updates; portfolio updates take effect at an institution's first renewal, extension or new agreement on or after Jan. 1, 2028. Vendor claims: 'more choice, more flexibility, more value'. Why now: announced the day before the last Microsoft run but not captured; it sets the licensing route by which Copilot reaches schools.
Limitations & uncertainty
Vendor claim; no prices disclosed on the page; details deferred to Microsoft Learn FAQ and account teams; impact depends on each institution's contract dates; applies globally, not only US.
- Source
- Claude Help Center (Anthropic)
- Published
- Undated; page states updated over two weeks before Nov. 3, 2026
Anthropic documents a free district-managed Claude for Teachers organization with K-12 DPA, SSO, SCIM and blocking of personal accounts on district domains
Per Anthropic, districts can now create a free Claude Enterprise organization for verified K-12 staff: an authorized administrator accepts K-12 terms, a public-sector addendum and a DPA once for all staff, verifies the domain, requires SSO with JIT or SCIM provisioning, blocks personal Claude accounts on the domain and migrates existing teacher accounts within 30 days. Connectors are off by default. The offer covers educators only, not students.
Limitations & uncertainty
Vendor claims; terms not reviewed here; free only for the first year, so renewal cost is unknown.
- Source
- arXiv
- Published
- Submitted November 4, 2026 (arXiv:2611.06762v1, 16:40:51 UTC)
arXiv preprint: state K-12 AI guidance converges on privacy and integrity but is thin on capacity and accountability
A Nov. 4, 2026 arXiv preprint (cs.CY; under review) analyzes publicly available AI-usage guidance from 35 U.S. states and Puerto Rico. The authors find that guidance priorities converge on topics anchored in laws such as data privacy and academic integrity, each addressed in 97% of documents. Most documents mention equity of access (86%) and algorithmic bias (97%), but fewer discuss specific marginalized populations. There is a common framing of urgency around AI integration to prepare students for workforce readiness (86%) and promote loosely defined personalized learning (86%), which is not matched by specific guidance for capacity-building, actionability, or accountability. State priorities showed no relationship to state demographics, capacity, or partisanship. Instead, the authors identify interdependencies through shared citations to organizations such as TeachAI and Code.org, with some states reproducing TeachAI’s model guidance near-verbatim, and peer citations concentrating on early states such as Oregon and Washington. No state includes parents as contributors on these documents, and only five include students. This paper was submitted two days before the last completed k12 run and was not in the archive; it is included now because it independently scrutinizes the quality of state K-12 AI guidance just as states are launching multi-state AI projects and districts are writing local rules.
Limitations & uncertainty
This is an unreviewed preprint (authors mark it “Under Review”). Only the arXiv abstract page was inspected; additional tables and methods in the 29-page PDF are not cited here. The corpus is publicly available guidance, which may omit informal memos or unpublished drafts. “More than half of U.S. states have published formal guidelines” is the authors’ broader context statement; the analyzed set is 35 states plus Puerto Rico. Percentages describe document mention rates, not implementation fidelity or student outcomes. The abstract does not name which states reproduce TeachAI language.
EdWeek Research Center: many educators lack deepfake policies or recent training as AI image abuse outpaces school response
A nationwide EdWeek Research Center survey of 852 educators conducted in summer 2026 finds one in three do not know if their school or district has a policy for AI-generated fake images and videos, 22% say there is no policy, 2% say one is being considered, and 13% do not know what a deepfake is. Nine in 10 educators reported no training in the last two years on detecting or responding to deepfakes. Experts quoted urge explicit policy language, a unified incident-response plan, trauma-informed victim support, and plain-language student notice; surveyed responses most often include informing parents (56%), suspension (48%), expulsion (25%) and restorative practices (33%).
Limitations & uncertainty
Self-report survey from summer; does not measure actual deepfake incidence in respondents' schools; recommendations are expert interpretation.
Houston ISD AI Guidebook v3 (Nov. 4) sets teacher-directed AI below grade 6 and parental consent plus teacher permission for older students
Houston ISD's AI Guidebook Version 3, last updated November 4, 2026, allows students below 6th grade to use only district-approved generative AI in teacher-directed learning, while students in grade 6 and above need written parental consent and prior teacher permission for each approved educational task. Students must disclose AI assistance and may need to produce chat histories; the district says it does not promote AI detectors because of accuracy and bias limits. Classroom-approved tools listed are Google Gemini, Microsoft Copilot, Microsoft Reading Coach and Adobe Creative Cloud Suite; other student-facing tools require AI Support Team vetting. Staff must not enter PII into AI tools, must verify consent, and grades 3-5 generative AI chats are teacher-modeled only without individual student chat access.
Limitations & uncertainty
Policy text only; consent uptake and classroom adherence unknown; living document may change after Nov. 4.
Schools lean on fragile workarounds to hide AI Overviews after Google declines to offer an off switch in Search
WIRED reports Google declined to give schools a way to disable AI Overviews; from next year schools must block Search entirely to remove them, though Search, Gemini and Gemini Notebook can be enabled separately. District IT staff already use workarounds such as modified search URLs and browser extensions (Fairfax built its own), which could break if Google changes result delivery. LA Unified is reviewing implications. A former Google Search PM says maintaining a forked Search is hard; MagicSchool's CEO notes Microsoft offers an extension to hide Bing summaries. USC's Stephen Aguilar expects under-resourced districts to stay with Google. A cited Washington University study found about 42 percent of AI Overviews fully verifiable from their citations.
Limitations & uncertainty
Google did not respond to WIRED; rollout specifics are pending; district examples are few; study figures are secondhand.
Accelerate / SETA-ED: nine-school quasi-experiment finds no detectable effect of EnlightenAI on AP U.S. History writing or exam scores, as students almost never revised
A 2025-26 quasi-experimental study across nine Uncommon Schools high schools (five using EnlightenAI, four comparison) found no detectable effect of access to the AI writing-feedback tool on spring writing scores or AP U.S. History exam scores. Nearly all students (328 of 329) logged in and submitted 4,846 assignments, but no student completed more than one of the 6 to 8 intended revision cycles and two-thirds completed none, so the study tested access rather than the revision-based model.
Limitations & uncertainty
Quasi-experimental with only nine schools, so the comparison has few effective data points; the source says it cannot distinguish a tool that does not improve writing from one not used as designed. Summary page of an independent evaluation published by the funder; the full report was not opened. Single charter network and course.
Google will switch Gemini and Search off by default for K-12 student accounts, but districts that keep AI off also lose Search
Google told districts in a Nov. 2 virtual meeting and a Nov. 3 statement that Gemini and Google Search will be off by default for K-12 student accounts until an administrator turns them on, with settings that can differ by grade; changes start early next year. Because Search and AI are integrated, traditional Search without AI summaries will not be offered, so groups barred from AI lose Search. Common Sense Media's Yvette Renteria warned of lost source-vetting practice and inequity for small districts lacking staff for workarounds such as filter vendors or Fairfax's Chrome extension hiding AI Overviews. Sens. Blumenthal and Hawley had set a Nov. 2 deadline for Google to answer nine questions; Google's statement did not address them.
Limitations & uncertainty
Rollout details are not final; changes start early next year; Guided Learning in Search is only being piloted; critics' and Google's claims are not independently tested.
Sweet Home Central (NY) builds a Claude-based after-school bus routing tool; superintendent says six-hour route builds drop to minutes, before rollout
A suburban Buffalo district of about 3,800 students built its own after-school bus routing tool on Claude with two regional partners and planned to roll it out the week after publication. The superintendent says routing a run by hand took about six hours and forced two-week advance sign-ups; he says the tool cuts this to minutes and enables same-morning sign-up. The district next plans "digital coworkers" for one administrative department.
Limitations & uncertainty
Self-reported, pre-rollout claims from one small district. "Minutes" is not quantified; no error rate or safety review of generated routes is reported.
Google announces Guided Learning in Classroom and a Search pilot, and off-by-default AI controls for K-12 student accounts
Google's VP of Education Chris Phillips announced that Guided Learning, which coaches students step by step instead of giving final answers, is globally available in Google Classroom starting Nov. 3 and will be piloted in Search in select U.S. schools later in November. Gemini, Gemini Notebook, Gemini in Classroom and Google Search will be entirely off for K-12 student accounts until administrators turn them on, with more specifics promised in coming months. Google states its research shows Guided Learning leads to measurable learning gains, but the post provides no study details here.
Limitations & uncertainty
Vendor claim; does not mention that opting out of AI removes Search; learning-gain research not detailed.
After backlash, Google says schools can enable Search alone without Gemini, but AI summaries stay in Search
NBC reports that after backlash to its Nov. 3 announcement, Google said schools could enable Search alone without turning on other Gemini products, but Search would still include AI features such as AI-generated summaries. Google said AI Mode and AI Overviews will offer step-by-step guided learning so students do not get quick answers. AFT president Randi Weingarten said Google prioritized business interests over children's education; Fairplay's Josh Golin said Google is trying to strong-arm schools and urged privacy-protective search alternatives. Until Tuesday, Gemini was automatically activated for schools using Google Classroom.
Limitations & uncertainty
Clarification came via social media; detailed admin guidance is pending; guided learning in AI Overviews is a vendor claim.
Stream editions
Each edition carries its own synthesis and evidence ledger.
November 10, 20261 edition
November 9, 20261 edition
November 8, 20261 edition
November 7, 20261 edition
November 6, 20261 edition
November 5, 20261 edition
November 4, 20261 edition
November 3, 20261 edition