Stanford researchers are challenging the way schools think about artificial intelligence in the classroom by arguing that educators should start with what students actually need to learn, not with what AI tools can do. A new report from Stanford University's AI Hub for Education identifies five critical learning experiences that research links to long-term student success, then asks whether AI can help schools deliver them at scale. The findings reveal a significant gap between what AI could accomplish and what's actually being built and deployed in schools today.

What Learning Experiences Actually Matter for Student Success?

Chris Agnew, managing director of Stanford's AI Hub for Education, and his colleagues identified five key learning experiences that research consistently shows matter for student development. These experiences form the foundation of their framework for thinking about AI's role in education.

• Personalized Instruction: Students receive teaching tailored to their individual learning level and pace, rather than one-size-fits-all lessons designed for an entire grade level.
• Real-World Learning: Students engage with authentic problems and contexts beyond textbooks, connecting academic knowledge to practical applications in their communities.
• Student Agency: Students have meaningful choice and control over what they learn and how they demonstrate their understanding.
• Enriching Discussions: Students participate in substantive conversations that develop critical thinking and expose them to diverse perspectives.
• Strong, Supportive Relationships with Adults: Students have consistent access to mentors and teachers who know them well and invest in their growth.

The challenge, according to Agnew's research, is that schools today struggle to provide these experiences at scale. Rigid scheduling, inflexible staffing structures, narrow accountability systems, and insufficient teacher training all create barriers to delivering the kind of learning that research shows matters most.

How Could AI Help Schools Overcome These Barriers?

Rather than asking what AI tools can do, Agnew's team flipped the question: if these five learning experiences are what matter, what's preventing schools from providing them today, and can AI help remove those obstacles? The analysis revealed several concrete ways AI could reshape how schools operate.

One major barrier is how students are grouped. Currently, schools organize students primarily by age, not by learning level or individual need. AI could synthesize assessment data with staffing and room constraints to enable much more dynamic grouping, allowing students to move between groups as they demonstrate mastery instead of staying locked into a fixed schedule set months in advance.

Assessment is another area where AI could make a significant difference. Today, schools measure a narrow slice of academic knowledge infrequently, which means results arrive too late for teachers to help struggling students. AI could support much more continuous, formative assessment and help teachers make sense of student data so they can intervene quickly. Importantly, AI could also provide visibility into a student's thinking process, how they revise their ideas, and how they collaborate, capturing skills like perseverance, creativity, and critical thinking that traditional testing rarely measures.

Professional development for teachers is a third opportunity. Simulation tools powered by AI could allow teachers to practice facilitating open discussions or responding to disengaged students, skills that traditional professional development doesn't build well because they require repeated practice with feedback, not a one-off workshop.

Why Aren't Schools Building AI Tools That Transform Education?

Despite these possibilities, most AI products being sold to schools are designed to fit into existing systems rather than reimagine them. Agnew explained the market dynamics driving this gap.

"A lot of today's purchasing decisions reward small improvements in the short term over bigger changes in the long term. AI products are often built to fit into how schools already work because that's what makes them adoptable and what brings ed tech companies revenue. This creates a gap between what might be possible in the long run and what gets built and used today," said Chris Agnew.

Chris Agnew, Managing Director, AI Hub for Education at Stanford University

Agnew argues that state-level systems could play a crucial role in reshaping this market. Instead of individual teachers or schools making isolated purchasing decisions, states could identify promising uses of AI and direct public dollars toward them. That kind of public signal could tell the private market to build toward longer-term systems change rather than incremental improvements.

What Does the Research Actually Show About AI in K-12 Classrooms?

Stanford's AI Hub also published a comprehensive analysis of existing research on AI in K-12 education, examining studies through November 2025. The findings paint a more cautious picture than the potential outlined above.

One striking finding is how limited the evidence base still is. There is very little research on AI in early childhood or pre-kindergarten settings, and much of the existing research focuses on post-secondary education. Until recently, there was no high-quality causal research on AI use by K-12 students in the United States.

When researchers examined the strongest available studies, they found mixed results. Academic outcomes improved when students had access to AI, but when the AI was removed, results varied significantly. In some cases, outcomes got worse; in one case, they actually improved. AI can ease students' mental load and make them feel better about their learning, but it doesn't necessarily encourage deeper thinking.

The design of the tool matters enormously. Purposeful, curriculum-anchored AI that provides step-by-step instruction shows greater promise than open-ended use where students have unlimited freedom to explore.

Stanford published the first rigorous research on U.S. K-12 students and AI a couple of weeks before the report. The study tracked 355 first through fifth-grade students in five after-school programs and two schools across two school districts. The results were sobering: nearly half of the elementary students never used their AI literacy tutor, even when there was dedicated time in the schedule to do so. When researchers paired students with human tutors, use of the AI support increased, but not enough to improve reading achievement.

Results for teachers using AI have been more promising. AI shows potential for reducing time spent on routine tasks, and automated instructional feedback can improve both teaching quality and student outcomes. Interestingly, the research suggests AI may be most beneficial to less experienced or lower-rated educators, potentially helping narrow performance gaps.

How Are Regional Ecosystems Advancing AI in Education?

Beyond Stanford's research, education is emerging as a key sector for AI adoption in regional innovation hubs. In India, the states of Telangana and Andhra Pradesh are positioning themselves as leaders in AI-enabled learning alongside broader economic development.

According to a white paper titled "AI for ALL: Catalysing Jobs, Growth, and Opportunity," developed by Prosus in collaboration with India's Ministry of Electronics and Information Technology and Boston Consulting Group, AI-enabled learning, personalized tutoring, and technology-driven classroom transformation represent the next wave of education technology innovation in the region.

"India's AI opportunity will ultimately be defined by how effectively we deploy AI across sectors that impact millions of lives every day. Telangana and Andhra Pradesh together illustrate the breadth of this opportunity combining world-class technology capabilities, thriving startup and research ecosystems, growing manufacturing ambitions and large-scale public sector innovation," said Sehraj Singh.

Sehraj Singh, Managing Director for India and VP of Global Corporate Affairs at Prosus and Naspers Group

The roundtable discussion highlighted how AI is already driving transformation across both states in education alongside agriculture, healthcare, and manufacturing. India accounts for nearly 16% of the world's AI talent and is expected to contribute almost 20% of incremental global economic growth over the next fifteen years, making execution at scale critical to the country's AI leadership.

The convergence of Stanford's research insights and regional deployment efforts suggests that the future of AI in education depends less on the technology itself and more on how intentionally schools and policymakers design systems around what students actually need to learn.