Research, policy, and news for schools—selected for teachers and administrators, synthesized in plain language, and paired with the limits each source still leaves open.
Reviewed
Coverage
U.S. K–12 + Ohio watch
Edition
12 analyzed items
CURRENT SIGNALS
What is moving now.
This opening edition establishes a trustworthy baseline for the new school year: Ohio's policy requirements, the strongest recent evidence reviews, national guidance gaps, and the safety and procurement questions moving fastest around schools.
01
AI policy is becoming operational work: approved tools, staff use, protected data, training, and recurring review now matter as much as student academic-integrity language.
02
The research base is expanding faster than the causal evidence. Assisted task performance, independent learning, and durable transfer still need to be separated.
03
Novel products can move faster than privacy review and community trust. Procurement evidence must cover the instructional role, data path, human oversight, and evaluation plan before deployment.
THE REVIEWED BRIEFING
Find what matters to your role.
Search every analyzed field or narrow by audience, topic, source type, jurisdiction, and reviewed edition. Filter choices stay in the URL so you can share the same view.
12 of 12 items
Current edition
Launch briefing · current through August 2, 2026
12 analyzed items
Research & Evidence
2
Evidence synthesisInternational
Published
Stanford SCALE Initiative
Stanford review finds a fast-growing literature but only 20 strong causal studies
What happened
Stanford's SCALE Initiative screened 818 papers in its repository and identified 20 that met its standard for strong causal evidence. The review reports that assisted task performance often improved, independent performance was mixed, pedagogically scaffolded tools appeared more promising than direct-answer chatbots, and some educator-facing tools saved time or supported instruction.
Why it matters for educators
The review separates completing a task with AI from developing knowledge that transfers when the tool is removed. That distinction is directly relevant to assignment design, product pilots, research claims, and decisions about what outcome a school should measure.
What this does not establish
The review does not prove that AI is broadly beneficial or harmful. No qualifying student-facing causal study occurred in a U.S. K–12 classroom, most outcomes were short-term, and much of the wider repository consisted of preprints or non-impact research.
Teachers
Administrators
Research Evidence
Classroom Practice
Tools & Procurement
Questions and sources
Question for your team
Is a product claim measuring assisted performance, durable independent learning, or something else?
New handbook chapter proposes a durable test for evaluating school AI
What happened
Stanford researchers Cristina Barnard, Lily Fesler, and Susanna Loeb organized the evidence around student-facing, educator-facing, system-level, and assessment uses. They propose evaluating tools by whether they expand access to personalized instruction, complex problem-solving, meaningful collaboration, substantive discussion, and sustained educator relationships.
Why it matters for educators
The framework connects fast-changing products to stable learning goals. Teachers can use those goals when considering assignment design, while leaders can use them to frame pilots and procurement discussions around learning experiences rather than feature lists.
What this does not establish
This is an evidence synthesis and decision framework, not a new controlled trial or a validation of any particular product. The authors also describe a thin causal evidence base, especially for U.S. K–12 settings.
Teachers
Administrators
Research Evidence
Classroom Practice
Tools & Procurement
Questions and sources
Question for your team
Does the tool expand or reduce students' access to experiences known to support durable learning?
Early-adopter districts are coordinating AI use, mostly within existing school models
What happened
CRPE surveyed leaders in 45 early-adopter systems, interviewed participants from 14 districts, and reviewed public policies and plans. It classified 58 percent as System Improvers, 16 percent as System Changers, and about 7 percent as Reimaginers, while identifying strain around evaluation, family engagement, procurement, and learning from pilots.
Why it matters for educators
The study offers a useful comparison set for districts moving from isolated tools toward coherent governance and instructional goals. Its categories can help leaders name whether a pilot is improving an existing process or attempting a deeper redesign.
What this does not establish
The participating systems were selected as early adopters, and the 45-system sample is neither exhaustive nor nationally representative. The study does not establish causal effects on learning, teacher workload, equity, or long-term organizational change.
Administrators
Policy & Governance
Tools & Procurement
Research Evidence
Questions and sources
Question for your team
What educational problem is the district solving, and what evidence beyond tool usage will show progress?
Virginia policy review finds gaps around staff use and student data
What happened
Virginia Commonwealth University's Metropolitan Educational Research Consortium reviewed the PK–12 AI-policy landscape and analyzed policies in 10 Virginia districts. The university's summary reports that six had no policy explicitly mentioning AI, only two addressed teacher use, and most had not revised data-privacy systems specifically for AI use of student records.
Why it matters for educators
Policies centered only on student plagiarism may leave staff use, procurement, professional learning, and protected student information insufficiently addressed. The review gives school teams a concrete checklist of governance areas to compare with their own documents.
What this does not establish
The 10-district snapshot is not a statewide compliance audit or a national estimate. It does not evaluate how consistently written policies are implemented or whether any policy approach changes teaching quality, student learning, privacy outcomes, or staff behavior.
Teachers
Administrators
Policy & Governance
Tools & Procurement
Research Evidence
Questions and sources
Question for your team
Does local policy clearly cover staff use, student records, approved tools, training, and human review?
Gallup finds formal AI guidance remains uncommon for public-school teachers
What happened
In a probability-panel survey of 2,069 U.S. public-school teachers, 18 percent reported formal guidance from administrators on workplace AI use, 48 percent reported only informal guidance, and 34 percent reported none. Lack of guidance varied by task: 69 percent reported none for one-to-one instruction or tutoring and 58 percent for grading and feedback.
Why it matters for educators
Written, task-specific expectations may be especially important where AI interacts directly with students or influences evaluation of their work. The results also show administrators where teachers report the most ambiguity and where professional learning may need clearer boundaries.
What this does not establish
The survey records teacher reports rather than auditing district policies. It excludes private-school teachers and does not show that guidance itself causes better student outcomes, safer tool use, lower workload, or greater teacher confidence.
Teachers
Administrators
Research Evidence
Policy & Governance
Professional Learning
Questions and sources
Question for your team
Which high-consequence tasks still depend on informal norms rather than clear written guidance?
Ohio publishes subject-by-subject AI integration guidance
What happened
Ohio released guidance connecting AI to its Computer Science and Technology Standards and to English language arts, mathematics, science, social studies, fine arts, financial literacy, physical education, and world languages. The package includes a Student AI Usage Model for activities such as brainstorming, editing, and group work.
Why it matters for educators
The documents give Ohio teachers a standards-connected starting point for discussing AI use by task rather than relying only on a blanket allowed-or-banned rule. They also give leaders a shared vocabulary for professional learning and curriculum conversations.
What this does not establish
The guidance does not mandate classroom AI use or demonstrate that its examples improve learning outcomes. It does not replace a district's adopted policy, approved-tool process, privacy obligations, or a teacher's judgment about a particular learning objective.
Teachers
Administrators
Classroom Practice
Policy & Governance
AI Literacy
Questions and sources
Question for your team
Which suggested uses preserve the learning objective and make the student's own thinking visible?
AFT proposes a restrictive, human-centered school plan for the AI era
What happened
The American Federation of Teachers released a 10-point plan calling for no screens in pre-K–2 except for compelling needs, no student-facing AI in elementary school, educator supervision of other student-facing uses, and a ban on companion chatbots through age 16. It also calls for active learning, stronger safety and privacy standards, educator control, independent research, and a technology tax.
Why it matters for educators
The plan is a prominent educator-union position that may influence bargaining, local policy discussions, procurement expectations, and debates over developmental readiness. It makes clear where one major stakeholder group wants schools to draw firmer boundaries.
What this does not establish
The plan is not law, government guidance, or a peer-reviewed evidence review. Its recommendations are advocacy positions and should not be presented as a settled consensus among teachers, families, researchers, school leaders, or policymakers.
Teachers
Administrators
Safety & Wellbeing
Policy & Governance
Classroom Practice
Questions and sources
Question for your team
Which parts are evidence-backed practice, which are policy demands, and which fit local student needs?
FTC asks seven companies how AI companions affect children and teens
What happened
The Federal Trade Commission voted 3–0 to issue orders to seven companies seeking information about companion-chatbot safety testing, youth access, disclosures, monetization, character design, moderation, and handling of conversation data. The agency said it wants to understand how firms assess and mitigate potential effects on children and teens.
Why it matters for educators
Schools may need to distinguish instructional assistants from products designed to simulate friendship or confidant relationships when setting age, safety, disclosure, privacy, and escalation rules. The category raises different questions from ordinary classroom productivity tools.
What this does not establish
A Section 6(b) study is an information-gathering inquiry, not a law-enforcement action or a finding that a company violated the law. The inquiry itself does not establish either harm or safety, and the cited agency pages do not yet publish final study findings.
Teachers
Administrators
Safety & Wellbeing
Policy & Governance
Privacy & Data
Questions and sources
Question for your team
Does local guidance address companion-style systems separately from instructional tools?
Federal AI grant guidance is joined by a final discretionary-grant priority
What happened
The Education Department's 2025 letter says existing formula and discretionary funds may support AI-based instructional materials, high-impact tutoring, and advising when uses remain within governing laws and program rules. Its 2026 final priority allows the Secretary to emphasize AI in selected discretionary competitions and includes age-appropriate AI education, educator professional development, ethical design, and universal design for learning.
Why it matters for educators
Districts and partners considering grant-funded AI work need to distinguish allowable uses, a selectable funding priority, and the actual requirements of a specific competition. The documents may shape proposals, but they do not replace program-specific review.
What this does not establish
The documents do not mandate AI adoption, automatically award funding, override program statutes or civil-rights and privacy obligations, or create a single national K–12 safety standard. A competition must actually invoke the priority for it to affect scoring.
Teachers
Administrators
Policy & Governance
Tools & Procurement
Classroom Practice
Questions and sources
Question for your team
Does the relevant grant notice invoke this priority, and what program-specific requirements still govern the project?
REL Central maps early AI uses for school-turnaround work
What happened
REL Central reviewed 48 sources describing AI uses plausibly relevant to school turnaround, including predictive modeling, tutoring, curriculum, and professional development. Twenty-two sources described implementation and 26 examined outcomes; the latter included 22 causal designs, but results and settings varied.
Why it matters for educators
The scan helps leaders separate specific use cases and evidence types instead of treating AI for school turnaround as one intervention. It also provides a starting point for checking whether evidence comes from a comparable setting and measures the outcome a district values.
What this does not establish
This was a limited rapid scan rather than an exhaustive systematic review. Many studies were not turnaround interventions in U.S. K–12 schools, and the document expressly states that mentioning products or services is not federal endorsement.
Administrators
Research Evidence
Tools & Procurement
Learning & Assessment
Questions and sources
Question for your team
Is the cited evidence from a comparable setting, and did it measure sustained school improvement or only a narrower outcome?
Salamanca pauses a proposed humanoid AI classroom pilot
Vendor claim boundaryDescriptions of the closed system and its handling of student data came through district or vendor accounts in reporting; they were not independent privacy or safety findings.
What happened
Salamanca City Central School District had approved a nearly $60,000 Realbotix system that included a stationary humanoid and virtual assistant for high-school technology work. The district paused the pilot while working on enhanced student-data agreements and community engagement after concerns from the state education department, educators, and residents.
Why it matters for educators
The episode illustrates how privacy review, the proposed instructional role, vendor relationships, community trust, and cultural context can become material procurement issues before deployment. It also shows why a pilot's purpose must be clear before a novel form factor drives the conversation.
What this does not establish
The robot was not deployed, and the pause is not necessarily a permanent cancellation. District or vendor claims about a closed system and data handling were not independent safety findings, and one unusual purchase cannot establish a broader trend.
Teachers
Administrators
Tools & Procurement
Safety & Wellbeing
Policy & Governance
Questions and sources
Question for your team
What privacy, instructional-role, community, accessibility, total-cost, and evaluation questions should be resolved before approving an unusual AI pilot?
Ohio Revised Code section 3301.24 required traditional public school districts, community schools, and STEM schools to adopt an artificial-intelligence use policy by July 1, 2026. The Department's optional model covers AI literacy, stakeholder engagement, privacy and security, procurement, acceptable use, academic integrity, and recurring review.
Why it matters for educators
Local policy now shapes approved tools, staff and student expectations, assignment-level disclosure, vendor review, and how concerns are reported. Teachers need usable classroom guidance, while administrators need implementation structures behind the board-adopted text.
What this does not establish
The law does not require a district to adopt the state model verbatim. Publishing a policy also does not show how consistently a district has translated it into training, approved-tool decisions, classroom routines, or enforcement.
Teachers
Administrators
Policy & Governance
Tools & Procurement
Privacy & Data
Questions and sources
Question for your team
Has the board-adopted policy been translated into an approved-tool list and usable classroom guidance?
Items are chosen for practical relevance to U.S. K–12 educators, with a standing watch on Ohio policy and guidance. Inclusion is not endorsement.
02
Synthesis
Every item separates what happened, why it may matter, and what the evidence does not establish. Stakeholder positions and provider announcements are labeled as such.
03
Sources
Original research, statutes, agency documents, and first-party reports are prioritized. Supporting reporting is attached when it adds context or documents a local event.
04
Publication
Monitored sources can produce a private candidate report, but candidates never publish automatically. A reviewed, tracked edition is the only public content source.
This briefing is an informational synthesis, not legal, policy, procurement, safety, or institutional guidance. Local requirements and circumstances govern school decisions.