Table of Contents
Microsoft AI in Education Report 2026: 77% Got No Training
The Microsoft AI in education report 2026 found 92% of students use AI and 77% got no training. Every number that matters, and how to close the gap at home.
Ninety-two percent use it. Seventy-seven percent were never taught how. That is the entire Microsoft AI in education report 2026 in two numbers, and the gap between them is where parents can actually do something. The report, published June 24, 2026, surveyed 3,345 respondents across six countries, including Brazil. It is the third annual edition, which means the trend line is now long enough to say something: adoption keeps climbing and instruction keeps lagging.
Key Takeaways
- The report surveyed 3,345 respondents across the United States, United Kingdom, Australia, Brazil, Japan, and Saudi Arabia, published June 24, 2026.
- 92% of students and education leaders and 88% of educators have used AI for school-related purposes.
- 77% of students and 53% of educators lack formal AI training; 66% of educators and 52% of students want monthly or quarterly training.
- 87% of educators and leaders and 79% of students agree AI skills matter for students’ futures, while 41% of students and 42% of educators worry about academic integrity.
- Microsoft paired the report with five free tools, including a Study and Learn Agent designed to give practice “without completing work for students.”
Every number in the Microsoft AI in education report 2026
Microsoft’s June 24, 2026 announcement laid out the findings. Here they are with what each one implies.
| Finding | Number | What it means for a parent |
|---|---|---|
| Students and leaders who have used AI for school | 92% | Assume your child has used it, regardless of school policy |
| Educators who have used AI for school | 88% | Assume their teacher has too, mostly for planning |
| Students reporting no formal AI training | 77% | The instruction is happening at home by default |
| Educators lacking formal AI training | 53% | Half of teachers are self-taught on this |
| Educators wanting monthly or quarterly training | 66% | The demand is coming from teachers, not being imposed on them |
| Students wanting monthly or quarterly training | 52% | Kids know they are missing something |
| Leaders reporting their school is implementing or scaling AI | 58% | Rollouts are ahead of training |
| Increased AI use over the past year (leaders / educators / students) | 78% / 76% / 65% | The trend is still up on all three sides |
| Educators and leaders who say AI skills matter for students’ futures | 87% | Near-consensus on importance |
| Students who agree | 79% | Students agree slightly less than adults |
| Worried about academic integrity (students / educators) | 41% / 42% | Students worry about cheating almost as much as teachers do |
The row that surprises people is the last one. Students are nearly as concerned about academic integrity as educators are. That contradicts the assumption that kids are casually indifferent to cheating, and it is a useful opening for a conversation: your kid may already be worried about this.
The training gap, triangulated
One survey is a data point. Three surveys pointing the same direction is a finding, and three surveys do.
Microsoft: 77% of students report no formal AI training. Common Sense Media, surveying 1,017 US teens in August 2026, found only 27% said a teacher had ever discussed safe AI use with them, meaning 73% had not. The Digital Education Council’s global survey of 45,398 respondents across 35 countries in July 2026 found 88% of students and 77% of faculty use AI, with 57% of students lacking guidance for assessments.
Three independent surveys, three methodologies, same shape: near-universal use, instruction reaching roughly a quarter of students.
Teachers are in a similar position. Microsoft found 53% of educators lack formal training. An NPR/Ipsos poll in June 2026 found only about a third of teachers report formal guidelines at their school, alongside 54% saying AI harms critical thinking and 59% saying it erodes trust. Gallup’s survey with the Walton Family Foundation found teachers using AI weekly save about 5.9 hours a week, and separately that most receive no formal guidance.
So the picture is not “schools pushing AI on reluctant students.” It is everyone using it and almost nobody being taught, with two-thirds of teachers asking for training they have not received.
Why the gap matters more than the adoption rate
Adoption is not controllable at this point. Instruction is, and the evidence says the difference between helpful and harmful AI use is a matter of habit.
PISA 2025, released September 8, 2026, found 46% of students across OECD countries use AI chatbots weekly or more for learning, that students using AI for summarizing, drafting, and research scored lower in science, and that moderate use was associated with better outcomes than either no use or heavy use. Associations, not causes, as the OECD notes. But the pattern separates two behaviors: AI that produces your work, and AI that tests your understanding.
Common Sense Media’s data shows which behavior is winning. Among teen AI users, 63% get an answer from the tool: 25% use it as-is, 31% rewrite it to sound like them, 35% improve it with their own knowledge. Meanwhile 66% say it “helps them understand schoolwork, not just finish it faster,” and 38% say it reduces their own idea generation. Both things are happening in the same population.
The skills teens say they value are worth noting: 55% cite working well with others, 50% fact-checking and critical thinking, 48% reading complex information. Those are not the skills AI does for them, which suggests students have a reasonable instinct about what matters and no structure for building it.
What Microsoft shipped alongside the report
The five free tools announced the same day are the company’s answer to its own finding. Unit Plans in Teach builds “fully developed, standards-aligned plans in minutes.” Student AI Guidelines and Learning Groups let teachers “set clear expectations for responsible AI use.” Learning Zone provides “educator-paced, live classroom experiences” with real-time visibility. Copilot Notebooks turns class materials into “structured, interactive study guides.” And the Study and Learn Agent delivers “research-based learning” through interactive practice and feedback in Copilot Chat “without completing work for students.”
That last clause is the design response to the 77% finding: if students have not been taught to use AI well, build a tool that refuses to be used badly. Whether that works is untested for this product, and the closest comparison is discouraging. Khanmigo used the same principle and a randomized study found students engaged with it on about one-third of their days in the platform.
What to do at home
Run the weekly error hunt
Since the training is happening at your table, make it a five-minute habit. Once a week, have your child ask an AI something they already know the answer to, and find the mistake. There usually is one. That builds the checking reflex that 77% of students are not getting at school.
Teach the two-category rule
AI that helps you check your thinking is allowed. AI that produces the thing you submit is not. That maps directly onto the PISA finding and is simple enough for a nine-year-old to apply. Our guide on when AI homework tools help versus hurt has the longer version.
Use the integrity number as a conversation opener
Forty-one percent of students worry about academic integrity. Say that out loud: “A lot of kids are worried about where the line is with AI. Where do you think it is?” That question gets a real answer more often than a rule does.
Ask the school for its training plan, not its policy
Two-thirds of educators want monthly or quarterly training. Ask when your child’s teachers last received AI training and when students will. A policy without training is a document; training is what changes what happens in the room. Our checklist for reading your district’s AI policy covers what to look for.
What not to do
Do not treat 92% as a reason to give up on rules; moderate use beat heavy use in PISA’s data. Do not assume your kid’s teacher has been trained, because half have not. And do not read the adoption numbers as evidence AI is helping, because the same report shows 41% of students worried about integrity and PISA found lower science scores among students using AI to draft and summarize.
What to Watch For Over the Next 3 Months
- Week 4: The weekly error hunt has happened at least twice, and your child can state the two-category rule in their own words.
- Month 2 red flags: Assignments completed faster with less evidence of thinking; your child cannot explain a step in work they submitted; no teacher has mentioned AI expectations to them all term.
- Month 3 self-check: Ask your child what AI is bad at. A kid who can name two specific weaknesses has been trained. A kid who says “nothing” has been using it without instruction, which is the 77% case.
Frequently Asked Questions
How big was the Microsoft survey?
3,345 respondents across the United States, United Kingdom, Australia, Brazil, Japan, and Saudi Arabia, published June 24, 2026 as the third annual edition. That is a moderate sample spread across six countries, so country-level breakdowns should be read cautiously.
Does 92% mean my kid is using AI for school?
Probably. The 92% figure covers students and education leaders who have used AI for school-related purposes, and two other 2026 surveys found similar rates. The more actionable number is 77% reporting no formal training, which is where a parent can change something.
Are teachers trained on this?
About half are not. Microsoft found 53% of educators lack formal AI training, and 66% want monthly or quarterly training. An NPR/Ipsos poll in June 2026 found only about a third of teachers report formal guidelines at their school.
Do students want AI training?
Yes: 52% want monthly or quarterly training, and 79% agree AI skills matter for their futures. Teens also name working with others (55%), fact-checking and critical thinking (50%), and reading complex information (48%) as skills that matter, which are precisely the ones AI does not do for them.
Is AI use hurting grades?
The evidence is mixed and directional rather than causal. PISA 2025 found students using AI for summarizing, drafting, and research scored lower in science, while moderate users outperformed both non-users and heavy users. Those are associations across very different school systems, not proof of cause.
About the author
Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.
Sources
- Microsoft. (2026, June 24). “Microsoft’s new AI in Education Report highlights widespread adoption and increasing demand for support.” https://news.microsoft.com/source/2026/06/24/microsofts-new-ai-in-education-report-highlights-widespread-adoption-and-increasing-demand-for-support/
- Common Sense Media. (2026, August 18). Teens in the AI Era: Schoolwork and the Skills That Matter. https://www.commonsensemedia.org/research/teens-in-the-ai-era-schoolwork-and-skills-that-matter
- OECD. (2026, September 8). “PISA 2025: Students’ reading and mathematics performance declined sharply across the OECD.” https://www.oecd.org/en/about/news/press-releases/2026/09/pisa-2025-students-reading-and-mathematics-performance-declined-sharply-across-the-oecd.html
- Gallup / Walton Family Foundation. (2025, June 24). “Three in 10 Teachers Use AI Weekly, Saving Six Weeks a Year.” https://news.gallup.com/poll/691967/three-teachers-weekly-saving-six-weeks-year.aspx
- Gallup. “Most Teachers Receive No Formal Guidance on AI Use.” https://news.gallup.com/poll/710534/teachers-receive-no-formal-guidance.aspx
- OECD. (2026). PISA 2025 Results (Volume I). https://www.oecd.org/en/publications/pisa-2025-results-volume-i_73451bc5-en.html
- Esbenshade, L., et al. (2026). Generative AI in K-12 Classrooms: A Midyear Implementation Report. AmplifyLearn.AI, University of Washington. https://arxiv.org/abs/2605.16277
- Chalkbeat. (2026, August 25). “Students rarely engaged with Khan Academy’s AI-powered tutor Khanmigo, study finds.” https://www.chalkbeat.org/2026/08/25/ai-tutoring-students-khanmigo-khan-academy-engagement-study/
