How should you use this citation hub?
Use it as a verified lookup table — every figure below includes the number, a short claim, the publisher and year, and a link to the primary page or PDF. We exclude Almanac-branded roundups as primaries, undated listicles, and vendor surveys without transparent methodology. Soft links to related Almanac guides are for implementation only — never treat them as primary citations.
Cross-survey caveat (read this first)
Gallup 60%, RAND 53%, EdWeek61%, and TALIS US 43% measuredifferent constructs (any AI vs generative AI vs instructional use), populations (all teachers vs ELA/math/science only; teachers vs principals vs school leaders), and waves. Donot treat them as contradictions or subtract one from another. Always cite study + date + n.
What share of US teachers use AI in 2024–2026?
There is no single official adoption rate. Cite each primary with its population and wording. Gallup/Walton (all public K–12), RAND (subject teachers), EdWeek Research Center (classroom intensity waves), NCES (school-leader reports), and TALIS (lower secondary) answer related but distinct questions.
| Stat | Claim + primary source |
|---|---|
| 60% | of US public K–12 teachers used an AI tool for their work in 2024–25 (Gallup/Walton ATP, n=2,232; Mar–Apr 2025) (Gallup, Jun 2025) |
| 32% | use AI tools at least weekly across nine work tasks (Gallup, 2025) |
| 28% | use AI less frequently (monthly or less) across those tasks (Gallup, 2025) |
| 66% | of high school teachers used AI tools (vs lower rates in other levels) (Walton/Gallup PDF, 2025) |
| 69% | of early-career teachers used AI tools (Walton/Gallup PDF, 2025) |
| 53% | of ELA, math, and science teachers used AI for school in 2025 (up >15 pp) (RAND RR-A4180-1, 2025) |
| 18% / 15% | of K–12 teachers used AI for teaching / had tried AI at least once as of fall 2023 (RAND RR-A956-21, 2024) |
| 25% | of ELA/math/science teachers used AI for instructional planning or teaching in 2023–24 (RAND RR-A134-25) |
| ~58–60% | of principals reported using AI tools for their work in 2023–24 (RAND RR-A134-25) |
| ~40% / ~20% | ELA or science teachers vs general elementary or math teachers using AI (2023–24) (RAND RR-A134-25) |
| 61% | of teachers used AI “a little/some/a lot” in 2025 (up from 34% in 2023; 32% in 2024) (EdWeek RC via EdWeek, Jan 2026) |
| 21% | of teachers say they have never used AI and don’t plan to (2025 wave context) (EdWeek, Jan 2026) |
| 69% | of public school leaders have a favorable view of teachers using AI for job duties (13% unfavorable) (IES/NCES SPP, Feb 2025) |
| 73% | of public schools report at least a few teachers use AI for lesson plans, admin, materials, instruction, assessments, and/or grading (NCES SPP, Dec 2024) |
| 43% / 36% | US lower-secondary teachers used AI in their work vs OECD average ~36% (OECD TALIS 2024 US note) |
After you cite the data
When you're ready to turn verified time-saving findings into copy-paste K–12 workflows (FERPA/COPPA-aware), the US Teacher playbook is optional — never a substitute for citing Gallup, RAND, or NCES.
US Teacher playbook — $27How much time do teachers save with AI?
Gallup/Walton’s “AI dividend” is the clearest primary time-savings series for US public K–12 teachers. Figures are self-reported estimates among users — not observational time-use studies.
| Stat | Claim + primary source |
|---|---|
| 5.9 hrs/wk | estimated savings among teachers who use AI at least weekly (Gallup, Jun 2025) |
| ~6 weeks/yr | weekly users’ savings over a 37.4-week school year (Gallup, 2025) |
| 2.9 hrs/wk | estimated savings among monthly AI users (Walton/Gallup PDF, 2025) |
| 60–84% | of AI users say AI saves them time on the tasks they use it for; ≤7% say it takes more time (Gallup, 2025) |
| +26% | greater AI dividend in schools with an AI policy (2.3 vs 1.7 hours saved/week per teacher) (Walton/Gallup PDF, 2025) |
| 64% / 61% / 57% | of AI users see higher quality in modified materials / student-data insights / student feedback (Walton/Gallup PDF, 2025) |
| 57–74% | quality improvement range across tasks; ≤16% say quality decreases (Gallup, 2025) |
| ~50 hrs/wk | teachers’ average workweek (prior Gallup/Walton findings reiterated in the AI report) (Walton/Gallup PDF, 2025) |
| 72% | of school leaders agree integrating AI will make life easier for teachers (NCES SPP, Dec 2024) |
What do teachers use AI for?
Planning and materials dominate. Grading, tutoring, and data analysis trail — a pattern that appears in both Gallup/Walton (US K–12) and TALIS (US lower secondary).
| Stat | Claim + primary source |
|---|---|
| 37% | use AI at least monthly for preparing to teach (Gallup, 2025) |
| 33% | making worksheets or activities with AI at least monthly (Gallup, 2025) |
| 28% | modifying materials to meet student needs with AI at least monthly (Gallup, 2025) |
| 28% | doing administrative work with AI at least monthly (Walton/Gallup PDF, 2025) |
| 25% | making assessments with AI at least monthly (Walton/Gallup PDF, 2025) |
| 16% | using AI to grade at least monthly (Walton/Gallup PDF, 2025) |
| 14% | using AI for one-on-one instruction at least monthly (Walton/Gallup PDF, 2025) |
| 12% | using AI to analyze student data at least monthly (Walton/Gallup PDF, 2025) |
| 73% / 70% | of US TALIS AI users: summarize topics / generate lesson plans or activities (OECD TALIS 2024 US note) |
| 38% | of US TALIS AI users generate text for student feedback or parent communications (OECD TALIS 2024 US note) |
| 29% / 23% / 18% | least frequent US TALIS AI uses: practice scenarios / assess/mark work / review participation data (OECD TALIS 2024 US note) |
| 70% / 67% | of school leaders agree AI will help teachers stay relevant / be more effective (NCES SPP, Dec 2024) |
Training, PD, and formal guidance
Teacher-reported “no training” shares remain high even as school-leader surveys and EdWeek PD waves show rising provision. Treat school-reported PD and teacher-reported uptake as complementary, not interchangeable.
| Stat | Claim + primary source |
|---|---|
| 68% | of teachers had not engaged in any school/district-provided AI training in 2024–25 (Walton/Gallup PDF, 2025) |
| 18% | receive any formal guidance from administrators on how AI tools should be used (Feb–Mar 2026, n=2,069) (Gallup, 2026) |
| 34% | receive no guidance at all across 10 AI work tasks (Gallup, 2026) |
| 48% | receive only informal guidance on AI use (Gallup, 2026) |
| 69% | receive no guidance on using AI for one-on-one instruction/tutoring (Gallup, 2026) |
| 58% | receive no guidance on using AI for grading and student feedback (Gallup, 2026) |
| 67% / 26% | of public schools provided AI training to all or some staff; 26% trained all teachers (NCES SPP, Dec 2024) |
| 60% | of districts planned to have trained teachers about AI use by end of 2023–24 (RAND RR-A956-21, 2024) |
| 35% | of district leaders reported providing students with AI training (spring 2025) (RAND RR-A4180-1, 2025) |
| 50% | of teachers reported ≥1 PD session on using AI (Sep–Nov 2025; n=693) (EdWeek RC, Nov 2025) |
| 13%→50% | teacher AI PD trend: 13% (2023) → 29% (early 2024) → 42–43% (Oct 2024) → 50% (fall 2025) (EdWeek RC, 2025) |
| 43% | of teachers had ≥1 AI training session by Oct 2024 (n=1,135 educators / 731 teachers) (EdWeek RC, Oct 2024) |
| 70% / 27% | among US TALIS non-users: lack knowledge/skills / school lacks AI infrastructure (OECD TALIS 2024 US note) |
| 22% | of US teachers report high need for PD on AI for teaching/learning (top need) (OECD TALIS 2024 US note) |
| 38% / 29% | OECD avg: teachers with AI professional learning / reporting PD needs in AI (OECD TALIS 2024 expertise chapter) |
Policy, guidance, and academic integrity
Written AI policies remain uncommon. Principal-reported guidance, teacher-reported integrity policies, and school-leader written student policies sit at different levels — cite the respondent type.
| Stat | Claim + primary source |
|---|---|
| 19% | of teachers are employed at schools with an AI policy (Walton/Gallup PDF, 2025) |
| 70% vs 60% | AI use among teachers at schools with vs without an AI policy (Walton/Gallup PDF, 2025) |
| 40% / 28% | favor / oppose AI tools in K–12 schools (Walton/Gallup PDF, 2025) |
| 56% vs 16% | favor AI in schools among AI users vs non-users (Walton summary, 2025) |
| 45% | of principals reported school or district AI policies/guidance (2025) (RAND RR-A4180-1, 2025) |
| 34% | of teachers reported school/district AI policies related to academic integrity (RAND RR-A4180-1, 2025) |
| 18% | of principals reported guidance on AI use by staff/teachers/students (2023–24) (RAND RR-A134-25) |
| 31% | of public schools have a written policy on students’ AI use (NCES SPP, Dec 2024) |
| 18% / 30% / 20% | planning a student AI policy / neither policy nor known plans / don’t know (NCES SPP, Dec 2024) |
| 32% | of public schools have teachers using software to identify AI-generated student work (NCES SPP, Dec 2024) |
| ~90% / 42% | educators expect AI to change teaching at least somewhat / “a lot” in next 5 years (Dec 2024 EdWeek RC) (EdWeek, Feb 2026) |
| 15 competencies | UNESCO AI Competency Framework for Teachers — policy framework, not an adoption rate (UNESCO, 2024) |
Equity gaps in training and guidance
High-poverty schools and schools serving higher shares of students of color are less likely to report AI training or guidance in NCES and RAND data.
| Stat | Claim + primary source |
|---|---|
| 75% / 58% | schools with ≤25% vs >75% students of color that provided AI staff training (NCES SPP, Dec 2024) · (Child Trends, Aug 2025) |
| 13% / 25% | principal-reported AI guidance in highest- vs lowest-poverty schools (2023–24) (RAND RR-A134-25) |
| Directional | teachers and principals in higher-poverty schools less likely to report using AI tools (RAND RR-A134-25) |
| 65% / 58% / 57% | suburban vs urban vs town/rural teacher AI use (Walton/Gallup PDF, 2025) |
| 57% | of teachers agree AI will improve accessibility of learning materials (Walton/Gallup PDF, 2025) |
Student AI use, teaching, and concerns
Student adoption rose alongside teacher use. Instruction on how to use AI responsibly still lags — and anxiety about false cheating accusations is common.
| Stat | Claim + primary source |
|---|---|
| 54% | of students indicated they used AI for school in 2025 (up >15 pp) (RAND RR-A4180-1, 2025) |
| >80% | of students reported teachers did not explicitly teach them how to use AI for schoolwork (RAND RR-A4180-1, 2025) |
| ~50% | of students worry they will be falsely accused of using AI to cheat (RAND RR-A4180-1, 2025) |
| 61% / 22% | parents vs district leaders who say greater AI use will harm critical thinking (RAND RR-A4180-1, 2025) |
| 48% / 55% | middle / high schoolers who agree greater AI use will harm critical thinking (RAND RR-A4180-1, 2025) |
| 47% / 72% / 32% | public schools teaching some/all students about AI; high/secondary vs elementary (NCES SPP, Dec 2024) |
| 42% / 31% | leaders favorable / unfavorable toward students using AI for education (NCES SPP, Dec 2024) |
Teacher and educator sentiment
Early Pew attitudes (fall 2023) were cautious. Later EdWeek waves show expectation of change even when net impact views remain mixed.
| Stat | Claim + primary source |
|---|---|
| 25% | of public K–12 teachers say AI tools do more harm than good (fall 2023; n=2,531) (Pew, May 2024) |
| 32% / 6% / 35% | equal mix of benefit and harm / more good than harm / aren’t sure (Pew, May 2024) |
| 35% / 24% / 19% | high school / middle / elementary teachers saying more harm than good (Pew, May 2024) |
| ~33% / ~70% | OECD headline: around a third of teachers use AI; seven in ten worry it facilitates plagiarism/cheating (OECD TALIS 2024 overview) |
Soft related reading (not primary citations): AI tools for teachers 2026, AI automation for teachers, and how to use AI as a teacher.
Methodology notes & survey caveats
Gallup / Walton Family Foundation (ATP): Teaching for Tomorrow surveyed Mar 18–Apr 11, 2025 (n=2,232 US public K–12 teachers); Gallup article Jun 24, 2025. Full PDF: Walton Teaching for Tomorrow PDF. Guidance wave: Feb 9–Mar 2, 2026 (n=2,069) — Gallup formal guidance article.
RAND: Prefer HTML report pages — RR-A4180-1 (Sep 30, 2025; ~8,601 ELA/math/science teachers; student AYP samples), RR-A956-21 (fall 2023 / Apr 2024), RR-A134-25 (2023–24 AIRS / AMES).
NCES School Pulse Panel: School-leader reported (n=1,490 public schools; Dec 6–20, 2024 data; IES press Feb 19, 2025). Complements teacher self-report panels — not a substitute. IES press · NCES SPP results.
EdWeek Research Center: Stats typically released via EdWeek journalism with wave-specific n (e.g., fall 2025 PD n=693; Oct 2024 n=1,135 educators / 731 teachers). Topic hub: EdWeek AI technology. Medium-confidence rows where n was not restated in later articles were omitted from this hub.
Why 60% ≠ 53% ≠ 61% ≠ 43%: Different sample frames, definitions, dates, and instruments. Report each with n= and date — never collapse into a single “teacher AI adoption” percentage.
Exclusions: No Almanac-originated numbers as primaries; no undated listicles; no vendor surveys without transparent methodology; CDT/Edge Research 2025 figures omitted here (primary often Cloudflare-blocked; marked medium in inventory). UNESCO cited only as a competency framework — not an adoption rate.
Causal learning outcomes: Almost all figures here are self-reported use, time, quality, or attitudes. RCT / value-added evidence for teacher generative AI remains thin in these primary families.
Sources index
- Three in 10 Teachers Use AI Weekly, Saving Six Weeks a Year — Gallup (Walton Family Foundation partnership) (2025)
- Teaching for Tomorrow: Unlocking Six Weeks a Year with AI (PDF) — Walton Family Foundation / Gallup (2025)
- The AI Dividend — Walton summary — Walton Family Foundation (2025)
- Most Teachers Receive No Formal Guidance on How to Use AI — Gallup / Walton Family Foundation (2026)
- Artificial Intelligence Use by Students and Teachers, and AI Guidance (RR-A4180-1) — RAND Corporation (2025)
- Using Artificial Intelligence Tools in K–12 Classrooms (RR-A956-21) — RAND Corporation (2024)
- Artificial Intelligence Use in K–12 Education (RR-A134-25) — RAND Corporation (2025)
- School Pulse Panel — AI & cell phones press release — IES / NCES (2025)
- School Pulse Panel results hub — NCES (2024–25)
- Fewer AI trainings in schools serving students of color — Child Trends (citing NCES SPP) (2025)
- A quarter of U.S. teachers say AI tools do more harm than good — Pew Research Center (2024)
- More Teachers Are Using AI in Their Classrooms. Here's Why — Education Week (EdWeek Research Center) (2026)
- Teacher AI Training Is Rising Fast — But Still Has a Long Way to Go — Education Week (EdWeek Research Center) (2025)
- We're at a Disadvantage and Other Teacher Sentiments on AI — Education Week (EdWeek Research Center) (2024)
- What's Holding Educators Back From Adopting AI — Education Week (EdWeek Research Center) (2026)
- Results from TALIS 2024 — United States country note (PDF) — OECD (2025)
- Results from TALIS 2024 — full report hub — OECD (2025)
- AI Competency Framework for Teachers — UNESCO (2024)