AI in Education: How Schools and Universities Are Getting It Right
AI in education is the application of artificial intelligence tools and systems across the learning environment — in classroom instruction, curriculum design, teacher workflows, student support, and institutional governance. It spans K–12 schools and higher education institutions navigating the same fundamental tension: AI is already in students' hands, and the question facing every school is no longer whether to engage with it but how to do so in a way that protects what makes education human. Human Agency works with educational institutions on AI transformation programs designed around that tension — not around the technology alone.
Two fears, one decision
The honest conversation about AI in education starts with two fears that institutions are holding simultaneously, and that neither should try to dismiss.
The first is the fear of harm from AI. Real concerns exist about academic integrity, student data privacy, and the erosion of the human relationships that define meaningful learning. Students who outsource their reasoning to AI tools lose the practice of thinking through hard problems themselves. Teachers who over-rely on AI for assessment lose the diagnostic insight that comes from reading student work closely. These concerns are not theoretical.
The second is the fear of being left behind. Students are already using AI tools outside the classroom — to write, to research, to problem-solve — with or without guidance. Take-home assignments that have worked for years no longer offer any meaningful evaluation of student capability. Social media has already shown what happens when students use powerful digital tools without institutional support: unsupported use carries serious risks. Choosing not to engage doesn't protect students. It just means someone else shapes how they learn to use AI.
The answer — the one that works in practice — is to hold both fears at once. Use AI carefully and intentionally, while having an open conversation with staff and students about its role in the world they are preparing to enter. Schools that get this right protect what makes education human and prepare students for what is already here.
What AI actually changes in how schools operate
The most immediate impact of AI in educational institutions is on teacher capacity. According to RAND Corporation's 2024 State of the American Teacher survey of nearly 1,500 public K–12 teachers, 60% reported feeling burned out — with twice as many teachers experiencing frequent job-related stress compared to similar working professionals. Administrative load is a consistent driver: grading, documentation, lesson preparation, and communication consume time that would otherwise go to students.
AI tools that handle the administrative layer give teachers back hours that belong in the classroom. Lesson planning that used to take ninety minutes can take twenty. Routine parent communications can be drafted and personalized quickly. Assessment rubrics can be generated and adapted for different learner profiles without starting from scratch each time. The teachers who benefit most are often the ones most stretched — in under-resourced schools where administrative burden falls disproportionately on already-overtaxed staff.
For students, the change is about access. AI tutoring tools that adapt to individual learning pace provide support that used to depend entirely on teacher bandwidth or family resources. A student who doesn't understand a concept can work through it outside of class, at their own pace, without waiting for the next available moment of teacher time. A 2025 meta-analysis published in Humanities and Social Sciences Communications, reviewing 35 experimental studies involving 4,193 participants, found a moderately positive effect of AI tools on student learning outcomes, with meaningful gains in both cognitive and non-cognitive skills. The critical variable was how AI was used — as a complement to teacher instruction rather than a replacement for it.
The risk is not that AI makes teachers or students more efficient. The risk is that it makes the wrong things more efficient. A lesson-planning tool can produce a complete, standards-aligned plan while still reinforcing weak instruction. An AI tutor can keep students engaged while quietly removing the productive struggle students need to build real understanding. In both cases, the tool does not solve the educational problem by itself. It amplifies whatever instructional framework surrounds it. If the institution has a strong theory of learning, AI can support it. If it does not, AI can scale shallow pedagogy faster.
What governance looks like at the school level
AI governance in education is not primarily a technology problem. It is a community trust problem — and most institutions are behind on it.
According to an EdWeek Research Center survey of 924 educators conducted in late 2023, 79% of educators said their districts still did not have clear policies on the use of AI tools. The schools that responded to this gap by banning AI tools outright — typically in response to academic integrity concerns — have often produced the worst outcomes: students using AI without guidance, and teachers left without clear direction.
Effective governance in educational settings covers four areas. It defines acceptable use — for students, for teachers, and for the institution — in plain language that is actually communicated rather than buried in policy documents. It addresses data privacy, which is particularly acute in education: student data carries specific legal protections under FERPA and COPPA in the United States, and any AI tool processing student information must be evaluated against those frameworks before deployment. It establishes accountability — who makes AI tool decisions, who reviews those decisions, and what the process is when something goes wrong. And it creates a feedback mechanism so that policies evolve as the technology and the community's experience of it evolve.
The institutions doing this well are treating AI governance the same way they treat other aspects of school culture — as a conversation that involves students, families, and faculty, not just an administrative decision handed down from leadership. This mirrors the enterprise AI governance principle that good governance enables rather than restricts: the schools with the clearest policies are the ones where teachers feel most confident using AI, not the most constrained.
How teachers actually build AI into their practice
The teacher enablement question is where most AI-in-education initiatives succeed or fail. A school that deploys AI tools and sends teachers a tutorial video will see the same outcome as any enterprise that invests in technology without meaningful enablement: adoption by enthusiasts, resistance from everyone else, and usage data that suggests the investment isn't working.
RAND's analysis of district AI adoption found that the number of districts training teachers on AI more than doubled from 2023 to 2024 — from 23% to 48% of districts — but as of spring 2024, 7 in 10 teachers still had not received any training on how to use AI in the classroom. The gap between tool access and tool fluency is where adoption stalls.
What works is hands-on training built around the work teachers already do. When a math teacher learns AI-assisted lesson planning by using a tool to plan the actual lesson they're teaching next week — not a hypothetical — the learning sticks because it's immediately applicable. When a writing teacher explores how students might use AI to assist their drafts, then designs an assessment that accounts for that reality, the exercise changes how they think about teaching writing, not just what tool they have access to.
The schools that see durable adoption share a pattern: they identify AI champions within each department — teachers who are naturally curious and become the go-to resource for their peers — rather than relying on top-down rollout alone. Peer learning in educational settings outperforms formal training because teachers trust colleagues who share their daily context. This is the same dynamic Human Agency sees across industries when building AI literacy programs: behavior change is most durable when it happens through real work alongside trusted peers.
How Human Agency approaches AI transformation for education
Human Agency's education practice is built on the belief that education is the greatest force to expand and protect human agency — and that schools getting this moment right will shape who their students become.
"We don't start with the tech. We start with your people — who they are, how they work, and what they need. Then we build, train, and stay until it sticks." That applies as directly to a school faculty as it does to an enterprise workforce. Before recommending any tool or program, the team conducts extensive stakeholder interviews with teachers, administrators, and students to understand the specific institutional context. An independent school with high family engagement faces different questions than a large urban district with resource constraints.
The six services Human Agency offers for educational institutions — AI Readiness Assessment, Teacher Enablement, Curriculum Development, AI Culture Coaching, Governance Frameworks, and Leadership Workshops — are designed to work together rather than as standalone products. An institution that deploys AI tools without governance and enablement is unlikely to see adoption. An institution that builds governance without equipping teachers to use AI confidently is likely to produce anxiety rather than capability.
Zac Cogley, Director of AI Solutions at Human Agency and a former associate professor and AI ethicist, leads the education practice. The questions schools are navigating — about what AI does to learning, what it does to the development of judgment, and what responsible use looks like for young people — are not purely technical questions. They require cross-disciplinary thinking that bridges academic philosophy and practical implementation. The same commitment to expanding human agency that shapes every HA engagement applies with particular force in educational settings, where the stakes are a person's ability to think, choose, and shape their own life.
Frequently Asked Questions
How are schools and universities using AI in 2025 and 2026?
Schools and universities are using AI in three primary ways: to reduce administrative burden on teachers through lesson planning, grading assistance, and communication drafting; to provide personalized student support through AI tutoring tools that adapt to individual learning pace; and to build institutional capacity for AI literacy so that students graduate able to use AI tools responsibly and effectively. According to RAND, 48% of school districts reported training teachers on AI by fall 2024, more than double the rate from the previous year — but most institutions are still in early stages of systematic adoption.
What are the biggest risks of AI in education, and how do schools manage them?
The three most significant risks are academic integrity concerns, student data privacy, and the erosion of skills that require productive struggle to develop — critical thinking, writing, and sustained attention. Schools managing these risks effectively are redesigning assessments to require personal demonstration and reflection rather than banning AI outright, vetting all AI tools against FERPA and COPPA data privacy requirements before deployment, and building AI culture programs that give students the judgment to use AI well. Human Agency designs governance frameworks and AI culture coaching programs for educational institutions working through exactly these decisions.
What does an AI readiness assessment for a school look like?
An AI readiness assessment for an educational institution evaluates AI tools currently in use — including shadow use that isn't officially sanctioned — staff AI fluency across roles, existing policies and their gaps, data infrastructure and privacy compliance, and the highest-impact opportunities for AI to reduce administrative load and improve student outcomes. The output is a clear, prioritized roadmap, not a 100-page report. It includes an honest picture of where the institution stands today across technology, people, governance, and culture, and a 90-day plan for where to start. Human Agency's process includes conversations with teachers, administrators, and where appropriate students — because the view from the classroom is different from the view from the district office, and both matter.
How do schools get started with AI transformation without getting it wrong?
The most common mistake is starting with the technology — selecting tools before understanding what problems the school is actually trying to solve and what capacity exists to implement them well. The schools that get it right start with leadership alignment: a shared, honest conversation among administrators, faculty, and board about what the school believes AI should do in an educational setting and what it should not. From there, the fastest path to durable adoption is a focused pilot — one or two use cases, one or two departments, implemented well enough to generate real proof of value before scaling. Human Agency works with educational institutions through this process from readiness assessment to implementation to ongoing enablement.
