AI education market size statistics

The AI education market is valued at an estimated $9.58 billion (Precedence Research)

AI is already being used in education at scale.

The largest deployment segment is cloud, with 57% of the market share. The largest component segment is solution (72%), the largest technology segment is machine learning (64%), and the largest application segment is learning platform and virtual facilitators (47%).

Expect to see even more use cases in the future.

The AI education market will be worth $42.48 billion in 2030 (The Business Research Company)

That’s based on a projected compound annual growth rate of 41.5%, an increase from the 40.9% forcecast between 2025 and 2026.

The AI education market will be worth $136.79 billion by 2035 (Precedence Research)

That’s based on a market of $7.05 billion in 2025 with a compound annual growth rate of 34.52%. Precedence Research called that CAGR “remarkable.”

The US AI education market is worth $2.73 billion (Precedence Research)

North America is the largest sector of the global AI education market.

By 2035, the US AI education market will be worth $39.83 billion (Precedence Research)

That’s based on a CAGR of 34.8%, a little higher than the CAGR for the global AI education market.

K-12 student use of AI statistics

Over half of US students aged 13 to 17 have used AI for research purposes (Pew Research)

The top reason for using AI is to search for information (57%). This is followed by help with schoolwork (54%), fun or entertainment (47%), and summarizing articles, books, or videos.

15% of students admit they use ChatGPT without permission from their teacher (Walton Family Foundation)

It’s somewhat common for educational institutions to ban the use of AI tools. But that doesn’t stop many students from using them. There are presumably more students who use AI tools without permission who are not willing to admit it, too.

30% of students use AI daily (Microsoft)

This figure is likely to increase as AI tools become more powerful and commonplace.

96% of younger students think digital learning tools are fun (New Schools)

Engaging younger students is crucial. Digital learning tools, including those powered by AI, can be a great resource for improving student engagement.

Around 1 in 3 educational institutions offer students AI training (Amazon)

Of those institutions, just 19% incorporate AI skills training. That means just 5.89% of educational institutions offer AI skills training to students.

Differences in which students use AI tools

95% of students now use AI (HEPI)

AI use is now borderline universal among students. As many as 94% report using generative AI to help with assessed work.

Although relatively low, the proportion of students directly including AI-generated text in assessed work continues to increase. As of 2026, this figure stands at 12%, up from 8% in 2025 and 3% in 2024.

Share of university students directly including AI-generated text in assessed work. Source: HEPI.

Students in grades 11 to 12 are twice as likely to use ChatGPT as students in grades 7 to 8 (Pew Research)

A little under a quarter of students in grades 11 and 12 (24%) reported using ChatGPT. In grades 9 and 10, that figure is 17%; in grades 7 and 8, 12%. Note that these figures are for students who’ve heard of ChatGPT. Those who haven’t were excluded.

Just 1% of university students and faculty in China never use AI tools (Technology Review)

The pervasiveness of AI tools in China’s universities might be a reflection of the country’s high optimism toward AI in general. That’s interesting, considering multiple research projects have China lagging behind the US in terms of AI capacity, investment, deployment, etc.

Students who study 3+ hours per weeknight are 11% more likely to have used an AI chatbot (Quizlet)

61% of students who reported studying for one or two hours per weeknight said they’d used ChatGPT or a similar tool. 72% of students who study three or four hours per night said they’d used these tools.

Black students are 12% more likely to use generative AI to get help with their homework compared to White students (Common Sense Media)

59% of Black students use generative AI for homework help. 47% of White students do the same. Black students are also more likely to use generative AI to get advice on a personal issue (25% compared to 14%) and help them write code (17% compared to 7%).

Why students use AI

Around 1 in 3 students use AI tools to help them get started on work, the most popular use case (Microsoft)

AI can be a great tool for beating writer’s block or introducing yourself to a new topic.

Other popular use cases for AI among students are summarizing information (33%), getting answers more quickly than through other means (33%), receiving initial feedback on schoolwork (32%), helping with studying (30%), improving writing (28%), and making projects more visually appealing (25%).

51% of university students use AI to save time (HEPI)

This isn’t unique to students. Saving time is one of the most common reasons people in general use AI. One recent survey found the average worker saves 1 hour per day thanks to AI. 5% of workers said AI saves them between 3 and 4 hours per day.

50% of university students use AI to improve the quality of their work (HEPI)

Improving work is another common use case for AI among students. This also extends to the workplace. One study from June 2025 found that 76% of office workers think AI helps their careers. That figure rises to 87% among Gen Z office workers.

It also extends to teachers. For example, 74% of teachers say AI improves the quality of administrative work. Interestingly, research by Gallup finds that teachers who use AI more frequently are more likely to report that it improves the quality of their work.

44% of students use AI for research, the most common use case (Quizlet via PR Newswire)

AI tools can be excellent research aides. That use was followed by summarizing information (38% of students) and generating study guides or study materials (33%).

Around 2 in 3 teachers say students use generative AI to write assignments (Forbes)

According to teachers, students are most likely to use AI to generate essays or other written assignments. 31% of teachers say students use AI chatbots during exams (including using the voice feature), to translate text (29%), and to generate code (28%).

59% of teenagers use generative AI to search for information or facts (Common Sense Media)

That made research the most common reason teenage students use AI.

Here’s the full list:

Students interested in using AI effectively can benefit from structured learning paths, such as associate’s and bachelor’s degree programs in cybersecurity, which often teach foundational AI, data literacy, and ethical use of generative technologies.

67% of kids and teenagers are regular users of AI companions (Common Sense Media)

Broken down further, 69% of 15 to 17-year-olds use chatbots often/sometimes. That’s 4% higher than 12 to 14-year-olds (65%).

Boys (70%) are more likely to use AI than girls (64%).

Changes in the use of AI in education

The number of students who are regular users of genAI tools has increased by 17% since 2023 (AIPRM and Tyton Partners via Illinois Department of Education)

In 2023, a survey by Tyton Partners found that 27% of students regularly used generative AI tools. By 2025, that figure had increased to 44%.

94% of students use generative AI for university assessments (HEPI)

In 2024, a narrow majority (53%) of university students in the UK used genAI tools while completing their assessments. In 2025, that figure leaped to 88%, and has cotnued to rise to 94% in 2026.

65% of university students say the way they are assessed has changed a lot because of generative AI (HEPI)

Generative AI tools have revolutionized how schoolwork is assessed. AI tools can assess work themselves, but also necessitate different assessment procedures due to how they can be used by students to cheat.

133% more university students believe their institution’s staff are well-equipped to work with AI than in 2024 (HEPI)

In 2024, just 18% of university students thought their institution’s staff were well-equipped to work with AI tools. By 2025, that had increased to 42%.

Rules surrounding the use of AI in education

Teachers are 3.8x more likely to allow students to use ChatGPT than to catch them using it when it is forbidden (Walton Family Foundation)

38% of teachers allow their students to use ChatGPT, while 10% say they’ve caught their students using it when not allowed.

Just 13% of universities around the world have institutional guidance or policies regarding AI tools (UNESCO)

The lack of institutional guidance regarding AI extends to the workplace. One survey found that just one-quarter of workers have received training on how to apply AI at work.

And only 7% of schools have guidance or policies (UNESCO)

The lack of institutional guidance is even more stark in elementary, middle, and high schools. Structured programs such as AI degree programs can help students and educators gain the skills and knowledge needed to use AI effectively and responsibly.

Of the education institutions that did have guidance, 40% reported that guidance was informal (UNESCO)

By informal, we mean not written down. Instead, the institutional guidance or policies regarding AI use were communicated orally. Not having clear, written policies or guidance can lead to confusion.

5% of university students feel aware of their institution’s AI guidelines and believe they are comprehensive (Digital Education Council)

Even fewer institutions have guidelines that are comprehensive and effectively communicated to students.

Sources: UNESCO; Digital Education Council.

Approximately 3 in 4 local US policymakers demand more AI regulation (Stanford University)

There’s a widespread appetite for AI regulation. According to Pew Research, 58% of adults and 56% of AI experts are more worried that the US government won’t go far enough when regulating AI, rather than go too far.

All 50 states have introduced legislation on AI (NCSL)

Lawmakers are working to keep up with the rapidly evolving field of AI. DC, Puerto Rico, and the Virgin Islands also introduced legislation on AI in 2025.

How education professionals use AI

3 in 5 educators use AI tools (Forbes)

Interestingly, less experienced teachers are more likely to have adopted AI than teachers with more than 10 years of experience. This could be a reflection of age, as younger people are more likely to have less teaching experience and are more likely to use AI.

99% of education leaders have used AI at least once for school-related tasks (Microsoft)

While almost every education leader has tried AI tools at least once, far fewer use AI regularly.

And 53% of education leaders use AI daily, compared to 36% of teachers (Microsoft)

Education leaders are more likely to use AI tools regularly compared to teachers (especially compared to the general population). According to research by Exploding Topics, only 35.49% of people use AI tools daily.

Almost 1 in 3 educators use AI to brainstorm ideas, the most common use case (Microsoft)

AI tools can be a helpful way for professionals to explore different ideas.

Other popular use cases for educators are creating lesson plans (29% of educators use AI for this), simplifying complex topics (24%), identifying ways to help students improve (24%), and personalizing student feedback or instructions (23%).

35% of teachers want data on student progress and personalized instruction in an AI tool (New Schools)

These are the two most commonly cited features that teachers want in an AI tool. That student data should be immediately accessible and actionable. Instruction should be personalized based on each individual student’s skill level.

Other commonly cited features are engaging students with learning (30% of teachers want this), ease of use (30%), cost (26%), aligned with relevant standards (26%), and improving student learning outcomes (25%).

The fact that improving student learning outcomes is relatively low on the list of teacher priorities is interesting, especially as a failure to improve student outcomes is the most commonly cited reason for school districts stopping use of a new digital learning tool. Indeed, just 38% of teachers believe AI tools improve student learning outcomes a great deal.

65% of history and social studies teachers use digital learning tools to teach at least half their class, the highest percentage of any teacher (New Schools)

14% of history and social studies teachers use digital learning tools to teach all of their classes.

While just 42% of reading teachers say the same, the lowest percentage (New Schools)

Just 9% of reading teachers use digital learning tools to teach all of their classes.

For comparison, 53% of English and language teachers use digital learning tools to teach at least half of their classes. The figure is 50% for science teachers, 50% for math teachers, and 46% for special education teachers.

Over 1 in 3 teachers use AI to prepare to teach (Gallup)

Preparing to teach is the most common task for which teachers use AI. As we discuss below, 80% of teachers say AI saves them time when preparing to teach.

46% of teachers say they never use AI to prepare to teach. And just 3% use it every day to prepare to teach.

Meanwhile, 33% use AI at least once per month to make worksheets, and 28% to modify materials to meet student needs.

94% of teachers seek input on new digital learning tools from other teachers (New Schools)

Teachers are the most trusted resource for information on new digital learning tools among teachers, principals, and administrators. For comparison, 85% of teachers seek information from their school or district, 84% look up information online, 58% research social media, and 55% ask their students.

27% of teachers only use free AI tools (New Schools)

Infrequent users of AI are far more likely to use only free tools. Of teachers who use AI once or less per week, 40% use only free tools. Among teachers who use AI daily, 24% only use free tools.

Almost half of all teachers purchase AI tools with their own money (New Schools)

Unsurprisingly, teachers who use AI tools daily are almost twice as likely to do this (46%) as teachers who use AI tools once per week (28%).

It’s not just teachers who feel they need to buy their own AI tools. According to research by Exploding Topics, 29% of employees pay for their own AI, often without the knowledge of their boss.

87% of teachers use at least one AI tool to make their job easier (Amazon)

The most common use of AI tools by teachers is grading, according to this survey from Amazon. That’s followed by administrative tasks, then personalizing learning.

59% of educators would be interested in programs that “train the trainer” (Amazon)

In other words, a majority of educators are interested in learning how to train their students in AI skills. Computer science and data science degrees can equip educators with the knowledge and skills to confidently integrate AI into their classrooms.

51% of educators use AI for educational games, the most common use (Forbes)

AI educational games are becoming increasingly common, with a majority of educators already using them.

Other common uses of AI are adaptive learning platforms (43% of educators use them), automated feedback or grading (41%), chatbots for supporting students (35%), and intelligent tutoring systems (29%).

Which AI tools do students and teachers use?

Globally, approximately 9 in 10 students use ChatGPT (Digital Education Council)

That makes ChatGPT the most popular AI tool among students, at least in the 35 countries surveyed by the Digital Education Council. That survey included over 27,000 students.

56% of students have used an AI search engine, the most commonly used AI tool (Common Sense Media)

AI-powered search engines were followed by chatbots (51% of students have used one), image generators (34%), and video generators (22%).

One in five students uses AI search engines daily. Just 11% say the same about chatbots.

What people think of AI’s impact on education

71% of teachers and 65% of students believe ChatGPT will be an essential tool for their success in college and at work (Walton Family Foundation)

In part, this could be due to the rise in remote learning and work. According to an October 2025 survey, 60% of Gen Z workers who use AI tools talk to those tools as much or more than they talk to their coworkers. Almost half say their AI tools know them better than their boss.

Over half of public K-12 teachers believe that AI tools in the classroom will cause equal or greater harm than benefit (Pew Research)

25% of teachers thought AI tools would bring more harm than benefit, and 32% thought they would bring an equal amount of harm and benefit. Just 6% said AI tools would bring more benefit than harm.

54% of people trust AI won’t discriminate or show bias (Stanford University)

That’s a decrease from 2022, according to the Stanford Institute of Human-Centered AI.

48% of students worldwide worry they aren’t adequately prepared for an “AI-enabled workplace” (Digital Education Council)

As AI becomes more important in the workplace, students are rightfully becoming more concerned with being adequately prepared to utilize AI tools. According to Gallup, the use of AI at work nearly doubled between 2024 and 2025, with 40% of US employees now using AI a few times a year or more.

Daily AI use also doubled, from 4% to 8% of employees.

And 58% think they don’t have enough AI knowledge and skills (Digital Education Council)

A majority of students worry about their lack of AI expertise, which is perhaps unsurprising considering how few educational institutions provide clear guidance on AI use.

Around 4 in 5 students believe knowing how to use AI responsibly and efficiently is important for their future (Microsoft)

In part, this is because experts think AI will significantly change the landscape of work, making some jobs obsolete and others look very different. Research suggests that about half of all US workers are in industries that are vulnerable to disruption by AI in the short term.

72% of students aged 9-17 want support with learning AI (Microsoft)

It’s clear that the relative lack of AI guidance and education in many schools isn’t because of student disinterest.

42% of students want to use digital learning tools more often (New Schools)

Interestingly, that’s far less than what teachers, principals, or administrators would think.

Teachers, for example, estimated that 64% of students would want to use digital learning tools more. Principals put that figure at 73% and administrators 66%.

Over 4 in 5 educators think the lack of AI teaching resources is a problem (Amazon)

Educators, many of whom are worried about their own job security in the face of AI, are well-placed to understand how problematic a lack of AI education is for students.

There’s also the question of helping young people learn how to use AI properly and responsibly. “It’s irresponsible to not teach it,” one Texas superintendent told the New York Times in early 2025. “We have to. We are preparing kids for their future.”

93% of experts who trained educators on AI reported teachers having negative views of AI (RAND)

Negative views of AI are fairly common among educators. Of 14 trainers interviewed by RAND, 13 reported encountering at least one teacher with negative views of AI.

81% of K-12 computer science teachers think AI should be a part of foundational computer science education (Economic Graph)

However, less than half of K-12 CS teachers feel properly equipped to teach AI.

68% of university students believe using AI is essential in today’s world (HEPI)

AI is rapidly becoming more widespread. It was estimated that over 378 million people would use AI as of 2025, more than three times the number of global AI users in 2020.

7 in 10 educators agree that using ChatGPT for assignments is considered plagiarism (Neuroscience News)

There’s a difference between using AI to aid in schoolwork and using it to complete schoolwork for you. Teachers are worried about students becoming overreliant on AI and not actually learning the material they are supposed to be learning.

For example, one teacher in Massachusetts told the New York Times that about half his middle school students use AI problem-solvers, like Google Lens. While some only use the tools for homework, he’s caught others using AI to complete quizzes.

41% of teenagers think generative AI will have both positive and negative effects on their lives (Harvard)

This reflects a broader trend among people who are often uncertain about the effect AI will have on their lives. According to Pew Research, 38% of Americans are equally as excited as concerned about the increased use of AI in daily life.

73% of university students and researchers think AI is effective for research (Zendy)

39.6% of students and researchers (including academics and faculty) think AI is “very effective” for research.

Around 9 in 10 university students and researchers think AI tools should cost under $10 per month (Zendy)

AI costs vary widely. Personal accounts might be free or come with a relatively cheap monthly subscription.

But many tools have more costly professional or institutional subscriptions, which can be prohibitive for teachers, especially as many have to use their own money to buy AI tools, as we discuss below. University students in particular often have to be careful about their spending, so it’s unsurprising they would prefer cheaper AI tools.

65% of teachers are worried about plagiarism due to AI use (Forbes)

Plagiarism is the most common AI-related fear among teachers. It’s followed by a reduction in human interaction (62%), data privacy and security (52%), job displacement for teachers (30%), unequal access to AI resources (30%), and automation of manual tasks (23%).

94% of teachers think students need at least some education in AI (Forbes)

Just 1% of teachers believe there’s no need for AI education.

How different people view AI in education

Teachers of older students are almost twice as likely to view AI negatively as teachers of younger pupils (Pew Research)

35% of teachers believe AI will bring more harm than benefit when it comes to educating high school students. For middle school teachers, this figure is 24%; for elementary school teachers, 19%.

And they’re 2.24x more likely to be sure about their opinion about AI (Pew Research)

Just 21% of high school teachers said they were unsure about what they thought of AI being used in schools. Middle school teachers were also fairly certain, with 28% saying they were unsure. But 47% of elementary school teachers (almost half) reported being unsure of their thoughts on AI.

People in China are 2.13x more likely to be optimistic about AI compared to people in the US (Stanford University)

In the US, 39% of people believe AI-powered services and products offer more benefits than drawbacks. That’s similar to other Western nations like Canada (40%) and the Netherlands (36%). And it’s an increase in optimism from 2022.

But it’s a much smaller percentage than in China, where a stunning 83% of people believe AI brings more good than harm. Other countries in the region are similarly optimistic, like Indonesia (80%) and Thailand (77%).

Education leaders are 48% more likely than teachers to claim they know a lot about AI (Microsoft)

Almost three-quarters of education leaders (72%) said they know a lot about AI. That makes leaders far more confident than teachers (44%) and students (41%).

Education leaders are 1.5x as likely as teachers to believe that AI is integrated into their school or district’s curriculum (Microsoft)

There’s an interesting disconnect between how education leaders and teachers view AI integration in their schools and school districts. The vast majority of education leaders, 82%, believe AI is integrated in their area. But just 54% of teachers agree.

And they are 1.4x more likely to think AI literacy should be an essential part of basic education (Microsoft)

A little over three-quarters of education leaders (76%) agree that AI literacy is essential to a basic education, but only 54% of teachers agree.

Students are 25% less likely to be excited or optimistic about AI in education than teachers (Quizlet via PR Newswire)

50% of teachers are excited or optimistic about AI, but just 39% of students feel the same.

LGBTQ+ students are 11% more likely to say generative AI will have a “mostly negative” impact on them (Harvard)

A little over one-quarter (28%) of LGBTQ+ students think generative AI will have a “mostly negative” impact on their lives. That’s significantly more than among cisgender and/or straight students, of whom just 17% say the same.

Changes in how people view AI in education

70% of nations surveyed saw an increase in residents who think AI offers more benefits than drawbacks between 2022 and 2024 (Stanford University)

Researchers at the Stanford Institute for Human-Centered AI surveyed people in 26 nations. Of those, 18 nations saw an increase in positive views toward AI between 2022 and 2024.

AI optimism grew by 4% between 2022 and 2024 in the US (Stanford University)

In 2022, 35% of the US was optimistic about AI. Two years later, that figure increased to 39%. Optimism in the US lags behind other countries around the world, especially those in East Asia. But it’s similar to other Western nations like Canada.

But excitement about the increased use of AI in daily life declined by 8% between 2021 and 2025 (Pew Research)

Pew Research presents a slightly less optimistic picture of how Americans perceive AI. In 2021, 37% were more concerned than excited about the increased use of AI in daily life, and 18% were more excited than concerned. The remaining 45% were an equal mix of excited and concerned.

Fast forward to 2025, and half of Americans say they were more concerned than excited, 38% were an equal mix, and just 10% were more excited than concerned. Put another way: 57% think the societal risks of AI are high, while just 25% think the benefits of AI are high.

Benefits of AI in education

Over 4 in 5 teachers state that AI cuts time spent making worksheets, assignments, etc. (Gallup)

Producing educational materials is one area in which AI can save significant time. For context, 81% of teachers say AI saves them time when completing administrative work, 80% when preparing to teach, 79% when grading, 74% when modifying materials to meet the needs of students, 65% when supplementing teaching, 63% when analyzing patterns in student learning or achievement, and 60% when providing personal tutoring.

These figures are all for teachers who use AI tools for the relevant task at least once per month.

Squirrel AI improved student question accuracy rates by 15% (World Economic Forum)

In this case study, an AI tool helped students improve their accuracy when completing quizzes.

Students in Australia who used an AI chatbot saw their grades increase by 10% (Microsoft)

Similarly, an Australian study gave students access to Microsoft Copilot. The AI tool helped them increase their grades.

Students in Indiana who used AI saw a 40% reduction in the time it took to complete their work (Indiana University)

In a study conducted with Microsoft, students in Indiana were allowed to use AI to help with their schoolwork. Their grades increased by 10%, and the time it took them to complete their work decreased by 40%, a significant difference. However, students said they felt like the work they completed with AI was less their own work.

Around 9 in 10 teachers think generative AI will positively impact students’ future careers (Amazon)

Based on this study, teachers are significantly more optimistic about the impact of genAI on their students than the population in general.

16% of the total productive capacity that could be unlocked in the US by generative AI lies in the education, health, and social work sectors (Economic Graph)

Experts think generative AI will unlock substantial boosts to productivity. A 2024 analysis by EY, for example, estimated that productivity gains brought by generative AI will increase GDP by 2.5% by 2033, which the company says would be worth some $650 billion.

A significant proportion of that potential increase is set to come from the education sector.

34% of university students would put in more effort if their exams were at least partially assessed by AI (HEPI)

However, this is only slightly more than the 29% who would actually put in less effort.

Interestingly, research suggests that students are more likely to prefer AI graders when they are dissatisfied with the current system. This is especially true for students who receive low grades. Students receiving high grades, however, generally prefer to stick with human graders.

47% of students say AI has a positive impact on their learning (Quizlet)

Interestingly, this is far less than the proportion of employees who think AI has helped them at work. Perhaps that’s because “learning” is a little more personal, requiring more human input than many common work-related tasks.

Around 2 in 5 students and teachers believe AI has created a more equitable education system (Quizlet)

Bias when grading is a widespread problem in education. AI could help solve for implicit biases, which humans often struggle to overcome.

48% of teachers think AI tools will help students recover from pandemic learning loss (Quizlet)

COVID had a significant negative impact on student outcomes. In some parts of the country, students in 2025 were almost a full year behind typical pre-COVID reading levels.

According to the Annie E. Casey Foundation, the decline in learning throughout the pandemic “amounted to decades of lost progress.” Experts worry this loss could translate to significant economic losses, as entire cohorts of students graduate with a relatively weaker education than their predecessors. But AI has the potential to help recover from pandemic learning loss.

55% of teachers say AI improves educational outcomes (Forbes)

This figure may increase as AI tools become more capable, and more institutions provide adequate training and guidance on how to use them.

Almost 6 hours a week is saved by teachers who regularly use AI tools in their work (Gallup)

Here, “regularly” means at least once per week. The time saved is based on estimates provided by the teachers themselves. Over the course of a standard school year, this amounts to six weeks saved annually.

Barriers to AI in education

Around 3 in 5 students are concerned about the potential unfairness of AI evaluations (Digital Education Council)

While AI has the potential to help counteract human biases, AI tools can have biases of their own. Biases can occur during data collection, data labeling, model training, and deployment.

65% of school administrators say they stopped using a digital learning tool (New Schools)

Almost two-thirds of school districts stop using a digital learning tool after piloting or adopting it. The most common reason for this was that the tool failed to improve student learning outcomes.

Just 18% of school administrators think their district offers a lot of evidence-based information about AI tools (New Schools)

This is particularly problematic, as it places the onus of researching information on AI on teachers, who are already overworked and potentially unfamiliar with AI.

81% of educators cite time constraints as a major challenge preventing them from developing an AI training curriculum (Amazon)

Educators are often overworked as it is, and lack the time to properly develop an AI curriculum. They would likely benefit from a prepared curriculum or courses on how to prepare one. According to one survey by RAND, teachers work an average of 53 hours a week, seven more than the typical working adult.

Three-quarters of educators say they are held back by a lack of knowledge on how to develop or access an AI curriculum (Amazon)

Most teachers lack sufficient familiarity with AI to develop a curriculum. In order to properly teach their students how to use AI, they need support.

High-poverty school districts are 53% less likely to have AI-trained teachers (RAND)

67% of low-poverty districts have trained teachers on AI use, but just 39% of high-poverty districts have done the same.

78.6% of district leaders to a DIY approach to developing AI training for their school district (RAND)

11 of 14 district leaders interviewed by RAND did this. Why? Perhaps to ensure their district’s training was personalized. But there is also widespread suspicion of people professing to be experts in AI.

For example, one district leader told RAND that “there are people that are claiming to have the best practices and are making money hand over fist.” The leader said “a lot” of the AI workshops they attended left them feeling that it had been “a waste of time and money.”

36% of university students say their institution doesn’t encourage them to use AI (HEPI)

The same proportion (36%) agrees that their institution encourages them to use AI tools.

Around 2 in 5 university students currently receive AI skills training (HEPI)

This is concerningly low, considering how many people agree that AI skills are important – and are becoming more important – in the workplace.

53% of university students avoid using AI out of fear of being accused of cheating (HEPI)

That’s the most common reason students are put off using AI tools. It’s closely followed by hallucinations, which worry 51% of students.

65% of students think an overreliance on AI will negatively impact their academic performance (Digital Education Council)

That’s up from 50% the previous year.

It’s important for students to find a balance when using AI. Relying on AI tools too much can negatively impact learning, as students let the tool do all the work for them.

For example, a study from MIT found that students who used only their brains (rather than relying on external support from an AI tool or search engine) displayed “significantly different neural connectivity patterns.” As external help increased, brain connectivity decreased.

AI in higher education and research

19 in 20 students use AI tools (HEPI)

That’s a significant increase from the 66% of students who used AI tools in 2024, and continues to rise from 2025’s 92%. The most popular use of AI was generating text, followed by editing work and accessing university textbooks.

76% of students and researchers use or are exploring using AI (Zendy)

27% of students, academics, and researchers are just starting to explore AI. 26% use AI occasionally. 20% use it extensively.

92% of students globally use AI for their studies (Digital Education Council)

This survey included responses from over 27,000 students from across 35 different countries. The true percentage is likely lower, as the survey only included students who are connected to the internet.

Over half of students and researchers use AI for literature reviews (Zendy)

That’s the most common use of AI among university students, academics, and researchers. It’s followed by writing and editing (46%), data collection (44%), and data analysis (40%).

Just 26.1% of students and researchers aren’t using AI at all in their research (Zendy)

It’s clear that researchers recognize AI’s potential to help make research more efficient. Just one-quarter of university students and researchers refuse to use AI at all.

Future of AI & education

Around 2 in 3 teachers intend to implement ChatGPT in their teaching more often (Walton Family Foundation)

As AI tools become more capable, more teachers will likely integrate them into their teaching practices.

97% of teachers would like to use digital learning tools as much as they do now or more in the future (New Schools)

Digital learning tools have quickly become a key component of teaching. This is particularly true for young learners, who overwhelmingly find digital learning tools fun and engaging.

Educators expect a 52% increase in AI skills courses by 2028 (Amazon)

As AI becomes more important at both school and work, the availability of courses teaching AI skills will almost certainly increase.

An estimated 74% of school districts provided AI training for teachers in 2025 (RAND)

In 2024, just 48% of school districts had implemented training. Furthermore, that training is often one-off rather than consistent.

32% of education professionals say their institutions have investment plans for generative AI (Microsoft)

Those plans are for the next 12 months. Another 9% of professionals say their institutions are in the process of early testing and proofs of concept for generative AI tools.

All higher education institutions surveyed by Microsoft have leveraged or plan to leverage generative AI (Microsoft)

Not a single institution surveyed by Microsoft had no plans to leverage genAI.

60% of teachers believe AI will be used more widely in education, but won’t be central to it (Forbes)

Despite widespread agreement that AI will become more commonly used in education, most teachers still think humans will remain central.

Miscellaneous AI in education statistics

Almost half of all students aren’t permitted to use personal devices in the classroom (New Schools)

Personal devices are an easy way to access AI tools. However, they’re also a significant source of distractions that could negatively impact learning, so more schools are starting to ban them. The one million children in New York City’s public school system, for example, must put their phones into storage at the start of each school day.

By high school, 96% of students use a smartphone on a weekly basis (The Social Institute)

While phones are increasingly being banned in school, they’re practically ubiquitous among high school students. That means almost all high school students have access to AI tools.

The most common age for students to first get a smartphone is 11 (The Social Institute)

In total, 15.57% of students receive their first smartphone at 11, while 15.05% receive their first smartphone at 12, and 12.76% at 10. In other words, kids commonly receive smartphones before they are old enough to use AI tools like ChatGPT (which has a minimum age of 13).

Less than half of all parents have given their kids an AI safety talk (Common Sense Media)

The majority of kids use generative AI tools, but many parents have not had a conversation with their children about the risks AI poses. That’s concerning, considering major AI tools like ChatGPT technically require teenagers to receive parental permission in order to use them.

Black teenagers are 2.86x more likely than White students to be falsely accused of using generative AI for completing schoolwork (Common Sense Media)

Teachers were far more likely to falsely accuse Black students of using generative AI than White or Latino students. This reflects existing racial disparities in discipline; Black students face higher rates of disciplinary measures despite research showing they are not more likely to misbehave than students of other races.