Gallup: 59% of US Teachers Used AI Last Year, and 64% Say It Freed Time

About three in five US K-12 educators used AI tools for work last school year, and nearly two-thirds of those users say the technology gave them more time for the parts of teaching they care about most. That is the headline from a Walton Family Foundation-Gallup survey, conducted with The Mattering Institute and published on October 14, 2026. The AI findings are one section of a larger report on whether educators feel valued at work.

This post walks through the AI numbers, separates what Gallup measured from what it did not, and connects the result to the other school-AI stories we have covered, so you can judge how much weight the 64% deserves.

Quick reference: the numbers

All figures come from Gallup's release. The "64%" in the headline is the sum of the first two user rows, 13% plus 51%.

What counts as "using AI" in this survey?

Gallup's definition is broad. It counts chatbots, adaptive learning systems and "other interactive AI platforms" used for preparing to teach, making student learning materials or assessments, grading, providing feedback, or administrative tasks. That means a teacher who asked a chatbot for a quiz once and a teacher who runs an adaptive math platform every day both land in the 59%.

Because the definition is wide, the 59% tells you AI has reached the staffroom, not how deeply it is used. A headline adoption rate is a floor for awareness, not a measure of reliance. The report does not break out use by tool, subject or frequency in the release we could read, so questions such as which chatbot, how many hours a week, or whether students are exposed to the output stay open.

Age is the one split Gallup gives. Use is highest among educators aged 40 to 59, close to average under 40, and considerably below average at 60 and older. That runs against the stereotype that the youngest teachers lead adoption, though Gallup does not publish the exact rates by age in the release, so treat the pattern as directional.

The time-savings result, read carefully

Of educators who used AI, 13% said it gave them a lot more time and 51% somewhat more. Another 29% saw no effect. Just 7% said AI cost them time.

Two readings are tempting, and both overreach. The optimistic one says AI hands nearly two-thirds of teachers back time. The skeptical one says a third of users got nothing or lost time. The survey supports a narrower statement: most users report a net benefit, few report a net cost, and the typical benefit is modest. "Somewhat more time" is the largest group, and it is the vaguest answer on the scale.

Three caveats matter for anyone quoting this number:

1. It is self-reported. Gallup asked how AI affected the time educators had for what matters most. It did not time tasks or compare lesson-prep hours before and after.
2. It covers users only. The 64% is a share of the 59% who chose to use AI, not of all educators. Teachers who tried AI and stopped, or never tried, are not in that denominator. Across all educators the share reporting more time is far lower.
3. It frames time as freed for "what matters." The wording ties the benefit to the parts of the job educators value. It does not say AI improved learning outcomes, grading accuracy or student results.

This is the same pattern we saw in workplace surveys outside schools. In our look at the AI verification tax, vendor surveys reported time saved by AI being partly offset by checking and redoing the output. Gallup's education data asks a different question and does not measure rework, so it cannot confirm or refute a verification tax for teachers. It is worth asking the follow-up: how much of a teacher's freed time goes back into checking what the model produced?

Why administrative work is the interesting detail

The wider report finds that educators feel they matter most when working directly with students. Eighty-five percent say helping a student through something difficult makes them feel they matter a great deal, 74% say the same about seeing students perform well academically, and individual feedback and leading creative, hands-on activities follow at 57% and 50%. Administrative tasks are, in Gallup's words, "by far the least likely" to give educators a strong sense of mattering.

That sets up the logic of the AI finding. If paperwork, materials preparation and routine feedback are the least meaningful parts of the job, then a tool that trims them can be valuable even if it saves only a little time. It is a cleaner case for AI than "AI teaches the class." The survey does not test whether teachers actually reinvest the time in students, which is the claim the headline implies.

Recognition shows the same student-centered picture. Sixty-seven percent of educators say they receive recognition from students at least a few times a month, compared with half who say so for supervisors and 38% for parents. Yet only 35% say they matter a great deal at their school, with 47% saying somewhat and 19% not too much or not at all. Gallup calls this the mattering gap.

How this fits the rest of school-AI news

Classroom desks with a spark above the teacher's desk, illustrating Gallup teacher AI survey results

The survey lands in a month when schools are pulling in different directions on AI.

• Bans for students, tools for staff. New York City banned generative AI for students in pre-K through 8th grade for 2026-27, covering about 600,000 students, while allowing limited supervised pilots in high schools, as we reported in the NYC schools AI ban. Gallup's data is about educators using AI themselves, which is a separate question from student access. A district can restrict one and encourage the other.
• Vendors courting teachers. Anthropic made a free offer for K-12 staff, covered in Claude for Teachers. Adoption numbers like Gallup's are what that kind of program is chasing.
• Research on generated classroom tools. Google Research's work on generative UI for teacher simulations and Microsoft's StudentSim tutoring simulators show where the tooling is heading: not just chat, but purpose-built classroom software.
• Integrity tooling. AI watermarking and its effect on students and teachers covers the other side of the classroom, where teachers worry about student work rather than their own workload.

Read together, these say schools are not adopting or rejecting AI as a block. Teachers use it as an assistant, districts set rules for students separately, and vendors and researchers are building for both.

What this means for teachers, school leaders and builders

Teachers. If you are in the 41% who did not use AI, the survey is not a reason to feel behind. The median benefit reported is "somewhat more time," and most gains sit in preparation and administration. A sensible trial is narrow: pick one recurring task such as drafting a differentiated worksheet or a first pass of feedback comments, time it for two weeks with and without the tool, and keep the review step.

School leaders. Do not cite 64% as an efficiency gain across your staff. Ask what share of your own teachers use which tools, for which tasks, and what they check afterward. Pair any rollout with a policy on student data, since the same tools that draft materials can touch names, grades and learning plans.

Builders and edtech teams. The report points at the target: administrative and preparation work that educators rate as least meaningful. Products that cut that work while keeping teachers in control of the final call fit the data. Products that promise to replace the student-facing parts of teaching run against what educators say gives them a sense of mattering. If you build for classrooms, the loop engineering career guide for students and our AI curriculum guide for high school students show what learners are being taught to expect from these tools.

What we could not verify

We read Gallup's public release. We did not have the full Walton Family Foundation-Gallup report, The Mattering Gap Among K-12 Educators, which the release points to, so we cannot say which AI tools respondents used, how often, at what grade levels, or how answers differ by subject, school type or teacher experience. Gallup also did not publish hours saved. Any figure beyond those in the release should come from the full report.

Respondents come from the Gallup Panel, a probability-based panel recruited through address-based sampling and random-digit-dial phone interviews. Gallup states that the sample was weighted to match US K-12 staff by age, gender, education, region and race or ethnicity, using Census Bureau data. It also notes that question wording and practical difficulties in conducting surveys can introduce error beyond sampling error. Those are standard caveats, and they apply here too.

Bottom line

The defensible takeaway is modest but real: a majority of US K-12 educators already use AI, most of those users report at least some time benefit, and only a small group say it cost them time. What the survey cannot tell you is how much time, whether it improved teaching, or how much checking the output required. Treat the 64% as a sentiment measure and watch for studies that time the work.

Figures reflect Gallup's release of October 14, 2026 and may be revised in the full report.

Related reading

• Claude for Teachers: free for K-12 educators
• NYC schools ban generative AI for 600,000 students
• The AI verification tax: surveys on checking and rework
• Google Research: generative UI for teacher simulations
• Microsoft StudentSim: LLM student simulators for tutoring
• AI watermarking: impact on students and teachers
• AI curriculum for high school students