K-12

K-12 — the use of artificial intelligence in primary and secondary education, spanning AI literacy curricula, AI tutoring, teacher support, and safety considerations unique to younger learners. A 2026 PRISMA-guided systematic review of 197 studies (2016–2024) organizes the field around four opportunity domains — personalization, motivation, Assessment, and innovative teaching practices — while flagging persistent gaps in teacher training, ministerial ethical guidelines, and discipline balance.

Questions to Consider

• K-12 AI use demands stronger safety protections than higher education because younger learners are least equipped to detect unsafe or manipulative AI. How does that change what 'safe' must mean for tools used by children?
• Systematic reviews find the K-12 evidence base concentrates overwhelmingly on STEM, leaving the arts and humanities largely unstudied. What might a discipline-balanced K-12 AI evidence base reveal that STEM-only studies miss?
• Research finds teachers systematically overestimate their AI competency — a roughly 40% gap between self-report and performance — yet short training yields substantially higher integration. What might that gap mean for how schools prepare teachers?
• Most AI tools center dominant perspectives, yet 78% of teachers in one study found AI helpful for culturally relevant pedagogy. How can a tool that defaults to dominant views be steered to broaden representation rather than narrow it?
• Evidence suggests outsourcing AI work can harm learning, especially for older students — yet younger students may be less affected. Why might the learning penalty of AI use differ by age, and what does that imply for design?
• The K-12 review identifies eight under-explored research gaps, from the missing concrete teaching-unit examples to the underuse of robotics, wearables, and mobile communication. Which of these gaps most limits what schools can actually do today?
• Institutional AI policies often lack implementation guidance. What would it take to turn a K-12 AI policy document into concrete classroom practices and teacher training — and why might that translation so often fail?

Introduction

The evidence landscape

A 2026 PRISMA-guided systematic review (Marzano, 197 studies, 2016–2024) now anchors the K-12 evidence base, complementing earlier large syntheses such as the rapid review of GenAI and PreK-12 learners and the aggregated K-12 AI evidence base. Its four-question structure organizes the field into four opportunity domains — personalizing learning, motivating students, improving assessment, and enabling innovative, immersive teaching — with ChatGPT as the flagship case studied across a large share of the included articles.

The review's central finding is a sharp asymmetry: the promise of GAI in schools is broad and well documented, but the evidence of daily classroom practice is thin, concentrated in STEM, and skewed by a shortage of concrete teaching-unit examples and practical teacher training. This gap between aspiration and implementation is the defining feature of K-12 GenAI research.

A teacher-side synthesis adds the layer that tool-focused reviews leave out. A PRISMA review of 29 studies of teacher intervention in K-12 AI-based instruction finds that AI output — alerts, dashboards, automated scores, chatbot feedback — becomes teaching only through a cycle of monitoring, judgment, intervention and orchestration, and that the reported benefits for student performance, participation and confidence are conditional on the interpretability of the information, intervention timing, the level targeted, teachers' implementation feasibility and students' autonomy. Its corpus is concentrated in the same places as the wider K-12 evidence base: mathematics and science/STEM subjects, middle and high school grades, and United States settings.

Distinctive K-12 considerations

• Safety and guardrailing: K-12 AI use demands stronger Pedagogical Safety protections. EduZone, EduGuard, and tutor harm research specifically address child-safe AI interaction.
• AI literacy curricula: K-12 AI literacy modules develop age-appropriate AI understanding, while the vibe coding framework empowers teachers to create their own AI tools. A quasi-experimental study of an 18-hour, 5E-model critical media literacy program for fourth-grade Turkish students (Demir & Akar 2026) embedded generative AI (ChatGPT, Grammarly, Canva AI, Padlet) phase-by-phase as a pedagogical agent aligned to the Turkish Language and Social Studies curricula, producing large gains in media reading (+3.50), writing (+1.67), and total media literacy (+5.17, all p < .01) with between-group effect sizes of Cohen's d = 1.12–1.31, and qualitative growth across six domains of critical media literacy (digital self-protection and data privacy, purposeful and responsible media use, safe communication and boundary awareness, critical evaluation and misinformation awareness, online risk awareness, and media ethics/digital citizenship) — evidence that age-appropriate, discipline-embedded AI literacy curricula can yield measurable critical-media gains in the elementary grades.
• Tutoring at grade level: K-12 personalized companions and correct answer trap research examine tutoring effectiveness for younger students.
• AI tutoring's constraint is engagement, not capability. In 18 Tennessee middle schools, assignment to Khanmigo raised math achievement ~0.06–0.08 SD per school year — gains resembling Khan Academy practice without AI — because the median student messaged the tutor in only 17% of mistake sessions (Oreopoulos and Low, 2026).
• Developmental appropriateness: Child safety research and Special Education considerations address the needs of diverse K-12 populations.
• Age-tailored conversational AI: Vahedian Movahed & Martin (2025) deployed AMA, a topic-bounded chatbot with age-varied prompt tailoring, with 63 students (grades 1 and 6–8). Children showed broad openness to and high trust in the chatbot as an information source and actively tested its credibility with known-answer questions, while gaps in digital-safety awareness (some were willing to confide secrets) point to age-sensitive Scaffolding and explicit Privacy instruction.

Teacher-knowledge effects

• Teacher-knowledge effects. A cross-level study (46 teachers, 2,832 secondary students) found pedagogical AI knowledge — not technical AI knowledge — drove students' perceptions of AI for social good and their intention to learn AI, reinforcing a "pedagogy first, technology second" approach to K-12 AI teacher preparation.
• Continuous training as the binding constraint. The Marzano review identifies continuous, targeted teacher training on ICT as the single most persistent integration challenge, and it must cover technical, pedagogical, and ethical dimensions while adapting to rapidly evolving tools. Initial and ongoing training should be continuous, collaborative, game-based, transparency-focused, and grounded in professional development for advanced tools like ChatGPT. This extends to classroom collaboration-support AI: the Community Builder (CoBi) feasibility study in middle-school classrooms found that high-integrity teacher use depended on substantial professional learning, with under-prepared teachers drifting toward treating the tool as a performance monitor or into general AI-discussion rather than its intended collaborative focus.
• Unplugged Machine Learning literacy through familiar content. In an online professional-development program, Sung and Gunpinar (2026) had ten K–12 educators build a visible yes/no feature matrix for classifying 2D shapes, so the algorithm could be inspected rather than accepted as a black box, and Self-Efficacy toward AI use rose from 3.13 to 3.71.
• The competency gap. Research shows teachers systematically overestimate their AI competency — a roughly 40% gap between self-report and performance — yet brief training interventions (e.g., 4-hour prompting workshops) can yield substantially higher classroom AI integration. This makes teacher preparation and Teacher Education central to successful K-12 AI integration.

Four opportunity domains

The systematic review consolidates what GAI can do in schools into four recurring domains:

• Personalized learning. GAI tailors learning experiences to individual needs, pace, and readiness — the flagship benefit documented across companion systems and adaptive platforms.
• Readability and curriculum alignment. Bird (2026) fuses transformer text classification with computational-linguistics features to classify English literature by UK Key Stage (F1 0.996), packaged into a no-code web app so teachers can align reading materials to learners' levels — a concrete, discipline-relevant (English/literature) complement to the STEM-heavy K-12 evidence base.
• Motivation and engagement. Tools like ChatGPT make learning more engaging for younger students, supporting engagement and Motivation — though the review also warns this can tip into excessive reliance and technological dependency.

A 42-study meta-analysis qualifies that: the motivational effect was smallest in school settings (secondary g = 0.360, primary g = 0.311) against g = 0.886 at university, though those subgroups rest on only 9 and 4 studies (Fang et al. (2026)).

• Assessment. GAI improves Assessment methods, from formative Feedback to automated scoring, though bias in automated feedback remains a live concern. Razavi and Powers (2026) extend this to item calibration in K-5: across 5,170 math and reading items, GPT-4o's zero-shot difficulty ratings correlated moderately-to-strongly with Rasch-calibrated difficulties (r = 0.83 math, r = 0.81 reading) but were uneven across grades and no better than a grade-mean dummy regressor for grades K and 1 — a range-restriction limitation that matters for early-grade assessment. A feature-based approach (Large Language Models (LLMs)-extracted features into tree-based models) reached correlations up to r = 0.87, with gains most pronounced for early-grade items.
• A related question is which model a school runs at all: on NGSS-aligned middle-school (1,078 items) and high-school (1,150 items) sets, nine open-weight models from six labs mostly scored above 90% and model size did not predict accuracy — OSS-20b and Nemotron 3 Nano 30b a3 matched models several times larger — while item difficulty ran high and discrimination low, leaving the authors to ask whether the Benchmark is too easy (Schroeder et al. 2026).
• Innovative teaching practices. Immersive, game-based, and collaborative approaches make learning more engaging, with constructionist and project-based models showing particular promise. A concrete classroom-wide example is the Community Builder (CoBi), an AI collaboration-support system deployed across six middle-school classrooms that used speech recognition to visualize small-group "uplifting" discourse — evidence that real-time speech AI can work feasibly in noisy K-12 environments to support the relational dimension of learning.

Cultural relevance imperative

LLM-supported curriculum design shows promise for diversifying materials — in one study 78% of teachers found AI suggestions helpful for culturally relevant pedagogy. However, most AI tools center dominant perspectives, requiring deliberate equity-centered design to avoid widening opportunity gaps.

The caveat is that representation is not transformation: in an all-girls GenAI makerspace, girl-only grouping, women facilitators and feminine marketing acted only at the symbolic level, while cost barriers and English-only prompts persisted and only two of 44 generated images reflected the workshop's own climate-change topic (Liu et al. (2026)).

Policy-to-practice translation

Institutional GenAI policies largely lack implementation guidance. The systematic review underscores that ministerial guidelines addressing Ethics and Privacy remain underdeveloped, echoing the policy-to-practice translation problem: successful models transform policy documents into actionable teacher training modules, bridging the "what" (policy) and "how" (prompting instruction).

Scale and equity

K-12 AI deployment operates at societal scale — millions of students, compulsory education, and significant equity implications. Equity research examines whether AI widens or narrows opportunity gaps, including access and inclusive support for students with disabilities — one of the review's eight under-studied areas.

Raising the floor is not leveling the playing field: over a year of school AI instruction, 752 Hong Kong junior-secondary students gained confidence, motivation and ethical awareness, with low-agency late learners approaching parity, while objective AI-literacy gaps persisted and self-initiated learning predicted the largest cognitive gains (Liang et al. (2026)).

Eight research gaps

An innovative contribution of the Marzano review is its explicit enumeration of eight under-explored research gaps that shape a K-12 GenAI research agenda: (1) a lack of concrete examples of AI in teaching for constructing teaching units; (2) a need for practical, daily teacher training based on constant AI application; (3) skills in machine learning and a balance between disciplines, with STEM over-represented; (4) a lack of studies and experiments in Europe; (5) a need to define the teacher role and create specific pedagogical frameworks such as Technological Pedagogical Content Knowledge (TPACK) or AI4K12; (6) inclusivity and support for students with disabilities; (7) a connection with pedagogical theories and innovative methodologies; and (8) applications of emerging technologies such as wearables, robot control, and mobile communication.

Connections

K-12 connects to Pedagogical Safety (child protection), AI Literacy (student competency), Teaching (K-12 teacher transformation), Equity (access gaps), and Special Education (diverse learner needs). Its systematic-review evidence base links to Meta-Analysis and Systematic Review as a method, and its discipline-imbalance finding connects to STEM Education, Humanities and Social Science Education, and Early Childhood Education.

Implications for K-12 instructors

• Prioritize child-safe guardrailing. K-12 AI use demands stronger safety protections — use tools specifically built and vetted for children (EduZone, EduGuard) and be alert to the harms of unguarded tutors (tutor harm research).

• Build AI literacy age-appropriately. Use curricula and frameworks designed for younger learners (K-12 AI literacy modules), and don't assume students arrive with responsible-use knowledge.

• Close the teacher-competency gap. Teachers systematically overestimate their AI competency (~40% gap between self-report and performance), yet short training (e.g., 4-hour prompting workshops) yields substantially higher classroom AI integration — invest in your own preparation (Teacher AI Competency, Teacher Education).

• Keep training continuous and practice-grounded. The systematic review's core recommendation is that teacher training cannot be one-off — it must be continuous, collaborative, and based on constant daily AI application, not abstract theory.

• Diversify materials deliberately. AI is strong at culturally relevant material generation (78% of teachers find suggestions helpful), but tools default to dominant perspectives — actively use AI to broaden, not narrow, representation, and guard against opportunity gaps.

• Watch the learning-penalty risk. Evidence shows outsourcing AI work can harm learning, especially for older students (cognitive surrender, homework-outsourcing penalty) — design assignments that keep cognitive work in the loop and assess for durable understanding, not just completion.

• Distinguish a productive slowdown from a streak. A randomized field experiment with 6,000+ middle-school students found AI support slowed progress but raised next-attempt correctness after errors, while a three-correct-in-a-row rule lifted platform mastery by roughly 28.7 percentage points without improving a delayed test a week later (Oreopoulos et al. (2026)).

• A disengaged turn is not a stable state. Rho et al. (2026) coded 5,522 middle-school math utterances and found the highest-performing cluster recovered from mocking or wrong answers, while withholding was the one disengagement behavior that persisted rather than resolved.

• Modeled prompts can become templates students copy. Across two 120 min STEAM modules, twelve eighth-graders' GenAI logs were dominated by information requests (303 of 310) and exact copying of the instructor's instruction was their most frequent standpoint (47.1%), with evaluation and reflection at 0.5% of coded content and result verification never coded (Zhao and Li (2026)).

• Design teacher tools for judgment, not volume. More AI information is not better: real-time alerts expanded teachers' awareness but in some studies overloaded attention or pulled focus from their own observation, and a single teacher cannot act on more intervention targets than they can physically reach (Lee 2026). Tools that prioritize and explain what is worth acting on, and let teachers accept, revise, defer or reject recommendations, fit the work better than dashboards that surface everything — and the structural conditions (review time, class size, support personnel) are part of the intervention, not background logistics.

• Extend beyond STEM. The evidence base skews heavily toward STEM; educators and researchers in the arts and humanities have an outsized opportunity to generate the discipline-balanced evidence the field is missing.

• Translate policy into practice. Institutional GenAI policies often lack implementation guidance; turn policy into concrete, actionable classroom practices and training (policy-to-practice).

AI Adoption and Skepticism in K-12

• As AI enters schools, the staff it affects are increasingly urged to consult conversational AI about adoption. An algorithmic audit of ten frontier LLMs probed a rural K-12 AI-skeptic persona, finding eight of ten models acknowledged concerns before redirecting to AI-engagement framings — a model-dependent design outcome rather than an inevitable property of LLMs. Trust, skepticism, and human oversight are therefore central to how rural and other K-12 staff encounter AI.

Grade-band progression is asserted in these frameworks and tested in none of them. Kharbach et al. (2026) searched 339 records and included 50 K-12 AI literacy frameworks, and found no study that tracked the same students across more than one grade band. Sun and Chan's measurement invariance across years 10 to 12 supports consistent scoring rather than developmental growth. A K-12 plan can sequence content across grades, but that sequence is a design proposal rather than an evidence claim.

Common Misconceptions

• "Teachers know their own AI competence." Teachers systematically overestimate their AI competency — a roughly 40% distance between what self-report claims and what measurement shows. Self-Assessment is not a usable needs-analysis instrument at this level of divergence.
• "AI tools are culturally neutral in practice." Most AI tools center dominant perspectives, yet 78% of teachers in one study found AI helpful for culturally relevant pedagogy. Teachers are using these tools for a purpose the tools were not built to serve well.
• "The K-12 evidence base is one literature." A PRISMA-guided systematic review of 197 studies (2016-2024) organizes the field around four opportunity domains, which means findings under one domain are not evidence under another.
• Compulsory-school students acquire AI practice before AI curriculum. Bergdahl (2026) studied a Swedish school running an "AI on the schedule" pilot for year 7 students, in a country whose national curricula include AI training only from upper secondary upward. The instruction covered ethics, prompting, train-your-own-model, recommender systems and data collection across three phases — and still left data privacy (Mdn = 1) and algorithmic manipulation (Mdn = 2) the weakest registers, against governance at Mdn = 4. Where the curriculum is silent, a pilot can supply instruction; it does not thereby supply the outcome.

Connected Concepts

• Pedagogical Patterns — The level of the productive-failure, conceptual-change and oral-assessment studies
• Early Childhood Education — Early childhood and elementary AI education
• Pedagogical Safety — Child-safe guardrails for AI interactions with younger learners
• AI Literacy — Age-appropriate student understanding and responsible use of AI
• Teaching — Teacher transformation and AI integration in K-12 classrooms
• Equity — Whether AI widens or narrows K-12 opportunity gaps
• Special Education — Diverse learner needs in K-12 AI use
• Scaffolding — Support structures for AI-assisted K-12 learning
• Personalized Learning — Tailored instruction enabled by AI at the school level
• STEM Education — AI literacy and tools across K-12 science, tech, engineering, math
• Generative AI — The core technology behind K-12 AI tools and curricula
• Teacher Education — Training and upskilling teachers for classroom AI
• Sociocultural Learning — Social and cultural dimensions of K-12 AI learning
• Metacognition — Thinking about thinking; key to durable AI-era learning
• Educational Development — Ongoing professional development for teachers
• Learning Design — Designing lessons and materials with AI support
• Active Learning — Keeping cognitive work in the loop when using AI
• Computational Thinking — Foundational K-12 skill and AI curriculum target
• Intelligent Tutoring — AI tutoring systems for younger students
• Meta-Analysis and Systematic Review — The systematic-review method behind the K-12 evidence base
• Humanities and Social Science Education — Discipline-balance counterpart to STEM-heavy K-12 AI evidence
• Parents and Families
• Cognitive Surrender

Connected Articles

• Generative Artificial Intelligence (GAI) in Teaching and Learning Processes at the K-12 Level: A Systematic Review — Systematic review of generative AI in K-12 teaching and learning (Marzano 2026)

• Pedagogy first, technology second: Cross-level relationships between teacher professional knowledge and student learning in artificial intelligence (AI) education — Cross-level effects of teacher AI knowledge (TAIK, TPAIK) on student learning in secondary AI education (Shen et al. 2026)

• Does school-based AI education narrow readiness gaps? The role of prior agency-related learning — School AI education narrows psychological but not cognitive readiness gaps

• AI-supported problem-based learning for enhancing computational thinking — AI-assisted project-based learning and computational thinking

• Making AI tutoring productive: Evidence from a mastery-based math practice experiment — Making AI tutoring productive: mastery-based math practice

• One click away: AI tutoring with Khanmigo in a two-year school experiment — One Click Away: Khanmigo in a two-year school experiment

• EduZone: A Framework for Evaluating LLM Safety for K-12 Students and Teachers — EduZone: child-safe LLM interaction platform for K-12

• EduGuard: A Safe RAG-Based LLM Tutor for Programming Education — EduGuard: safe RAG-based LLM tutor guardrailing

• SafeTutors: Benchmarking Pedagogical Safety in AI Tutoring Systems — Research on harms of unguarded AI tutors for children

• Learning to Use AI for Learning: Teaching Responsible Use of AI Chatbot to K-12 Students Through an AI Literacy Module — K-12 AI literacy modules on prompting

• The Evidence Base on AI in K-12: A November 2026 Review — The K-12 AI evidence base aggregated at scale

• ECNUClaw: A Learner-Profiled Intelligent Study Companion Framework for K-12 Personalized Education — K-12 personalized AI companions

• A Guiding Framework for K-12 Teachers in Creating AI-powered Learning Technologies through Vibe Coding — Vibe-coding framework empowering teachers to build AI tools

• Young people, learning, and generative AI: A rapid literature review and implications for PreK-12 education — Rapid review: GenAI and PreK-12 learners (271 papers)

• Faster Completion, Less Learning: Generative AI Reduced Study Time on Math Problems and the Knowledge They Build — Cognitive surrender strongest for high schoolers, absent for Grade 5

• The generative AI learning penalty: Evidence from Chinese secondary education — The generative AI learning penalty: homework outsourcing harms learning

• Marked pedagogies: Examining linguistic biases in personalized automated writing feedback — Marked Pedagogies: bias in automated feedback for middle-school writing

• Code to Learn with Generative AI: A Theoretically Grounded Framework for Artifact Construction in Upper-Secondary Education — CtL-GenAI: constructionism framework for artifact construction

• Beyond "painting in pink": A critical case study of all-girls generative AI workshops in a European makerspace and implications for gender equity in computing — All-girls GenAI makerspace workshops and gender equity in computing

• Transforming clicks into critical thinking: An AI-based media literacy program for children — AI-based critical media literacy program for children

• Estimating item difficulty using large language models and tree-based machine learning algorithms — Estimating item difficulty using LLMs and tree-based ML

• A Feasibility and Implementation Integrity Study of the Community Builder (CoBi): An AI-based Collaboration Support System in K-12 Classrooms

• Ask Me Anything: Exploring Children's Attitudes Toward an Age-tailored AI-powered Chatbot

• Teacher intervention in K-12 AI-based instruction: a systematic review of processes, strategies, and effects — Teacher intervention in K-12 AI-based instruction: a systematic review

• A Benchmark for LLM's Understanding of Middle School and High School Science Topics — A Benchmark for LLM's Understanding of Middle School and High School Science Topics

• Effects of generative artificial intelligence on students' learning motivation: evidence from a meta-analysis — Meta-analysis of GenAI's effect on motivation, smallest in school settings (secondary g = 0.360, primary g = 0.311)

• Disengaged Conversational Behaviors and Learning in Student-AI Dialogues for Mathematical Problem Solving — Disengaged conversational turns in middle-school mathematics: withholding persists where mocking recovers

• Exploring Middle School Students' Interactions and Perceptions of Using Generative AI in STEAM Learning Environments — Modeled prompts can become templates students copy

• Bringing Artificial Intelligence Literacy Into Online Education: Machine-Learning Integration Through Geometry in K–12 Teacher Professional Development — Unplugged machine-learning literacy through familiar content

• K-12 AI literacy frameworks after generative AI: A scoping review - no K-12 framework follows the same students across grade bands; progression is proposed, not tested