Generative Artificial Intelligence (GAI) in Teaching and Learning Processes at the K-12 Level: A Systematic Review

Synthesis: This PRISMA-guided systematic review synthesizes 197 studies (2016–2024) on Generative AI in K-12 teaching and learning. It documents GAI's potential to personalize learning experiences, motivate students, improve Assessment methods, and enable innovative teaching practices — with ChatGPT as the flagship example — while surfacing persistent challenges: continuous teacher training on ICT, ministerial guidelines addressing Ethics and Privacy, and a notable shortage of concrete, discipline-balanced (beyond STEM) experiments and practical examples for daily classroom use. The review contributes a focused account of eight under-explored research gaps and proposes directions for future empirical work.

Key Findings

1. PRISMA-guided synthesis of 197 studies. The review examined articles published 2016–2024, selecting 197 relevant studies from 5 databases and 2 journals, using a Mentefacto Map to identify keywords and define replicable inclusion/exclusion criteria.
2. Four opportunity domains. GAI offers significant potential to personalize learning, motivate students, improve assessment methods, and introduce innovative, immersive teaching practices.
3. ChatGPT as the flagship case. Adoption of ChatGPT in educational environments exemplifies GAI's capacity to make learning more engaging and tailored to individual needs.
4. Persistent integration challenges. These center on the need for continuous, targeted teacher training on ICT and the development of ministerial guidelines that address ethical and Privacy concerns.
5. A discipline-imbalance gap. The literature concentrates on STEM, neglecting the arts/humanities and creative subjects where many experiments could be conducted.
6. Eight under-explored research gaps. These include a lack of concrete teaching-unit examples, practical daily training, balanced disciplines, European studies, defined teacher roles and pedagogical frameworks (e.g., Technological Pedagogical Content Knowledge (TPACK), AI4K12), inclusive support for students with disabilities, links to pedagogical theories, and applications of emerging technologies.

Background and Method

Following the launch of ChatGPT — the first user-friendly large language model — Generative AI has transformed human–machine interaction and promised to reshape K-12 education, but it raises complex ethical and knowledge-related issues for students and teachers, especially at a sensitive age level with more specific, individualized needs. The review argues that although empirical evidence attests to the effectiveness of emerging educational technologies, the literature lacks comprehensive guidance on access to resources, content creation, and methodologies that combine pedagogy with advanced tools.

Methodologically, the review followed the PRISMA method. It examined articles published between 2016 and 2024, selecting 197 relevant studies from 5 databases and 2 journals. The Mentefacto Map structured keyword identification and the definition of inclusion and exclusion criteria, ensuring a systematic and replicable approach.

Applications of GAI in K-12

The review's analysis organizes findings around four research questions (Q1–Q4). The applications of GAI cluster around personalizing learning experiences, motivating students, improving assessment methods, and introducing innovative and immersive teaching practices. ChatGPT and other GAI tools are explored across a large body of the included studies (Aktay 2022; Allam et al. 2023; Blake 2024; Chen et al. 2023; Chiu 2023; Cain 2024; Su and Yang 2023; among others), which examine benefits for engagement and tailored personalization alongside ethical, privacy, and data-quality concerns.

Opportunities and Benefits

GAI offers significant opportunities to personalize learning, motivate students, improve assessment, and introduce innovative and immersive teaching practices. ChatGPT adoption shows remarkable potential to transform educational environments, making learning more engaging and tailored to individual needs. Studies on creativity and educational innovation (e.g., Mishra and Henriksen 2024; Vartiainen et al. 2023) analyze how GAI supports creative and innovative pedagogy, while also flagging risks of technological dependency.

Challenges and Risks

Several critical issues require attention. The most prominent is the need for continuous teacher training on ICT, which must cover technical, pedagogical, and ethical aspects and be adapted to constantly evolving tools. The review also highlights the lack of ministerial guidelines addressing Ethics and Privacy concerns, the shortage of concrete studies and experiments (especially beyond STEM toward the arts/humanities), and limited practical examples of GAI use in daily teaching practice. Q4 responses emphasize concerns about excessive reliance on AI and its impact on fundamental human skills such as self-regulation and Creativity, echoing themes of over-reliance and misuse.

Integration Challenges (Q3)

Integrating GAI into K-12 practice poses both pedagogical and technical challenges. Pedagogically, specific tools for teaching AI must be developed and evaluated, with game-based and collaborative approaches making learning more engaging, and teachers must adapt AI to different learning styles. Technically, educational tools must be continuously evaluated and updated to remain relevant across different age groups and learning levels, and teachers must be trained in collaborative, experimental, and transparent approaches that reduce concerns and increase confidence. Initial and ongoing teacher training should therefore be continuous, collaborative, game-based, transparency-focused, and grounded in professional development for advanced tools like ChatGPT.

Research Gaps and Future Directions

An innovative contribution of the review is its identification of eight specific literature gaps: (1) lack of concrete examples of AI in teaching for constructing teaching units; (2) need for practical, daily teacher training based on constant AI application; (3) skills in machine learning and balance between disciplines, with STEM over-represented; (4) lack of studies and experiments in Europe; (5) 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) connection with pedagogical theories and innovative methodologies; and (8) applications of emerging technologies such as wearables, robot control, and mobile communication. Future research should develop comprehensive frameworks and specific training practices, address ethics and long-term impact on educational outcomes, and adopt collaborative, interdisciplinary approaches to ensure inclusive, ethical, and responsible integration.

What this means for practice

• Instructors. Build teaching units from concrete examples rather than integration strategies: the review's first identified gap is the lack of practical examples for constructing teaching units.
• Instructors. Adapt AI to different learning styles and evaluate each tool continuously, replacing one-off introductions with continuous, collaborative, game-based, transparency-focused training grounded in daily application.
• Administrators. Draft ministerial guidelines covering Ethics, Privacy, and data quality, which the review identifies alongside teacher training as the binding constraint on responsible GAI integration in schools.
• Administrators. Fund work beyond STEM: the reviewed literature overrepresents STEM and neglects the arts, humanities, and creative subjects, where many experiments remain to be run.
• Researchers. Work the eight identified gaps deliberately, including discipline-balanced designs, European contexts, defined teacher roles and frameworks such as Technological Pedagogical Content Knowledge (TPACK), and concrete support for students with disabilities, plus designs that protect the self-regulation and Creativity the review flags as at risk from excessive AI reliance.

Limitations

• The synthesis rests on 197 studies published 2016–2024, selected from 5 databases and 2 journals with inclusion criteria structured through a Mentefacto Map, and the author presents GAI in K-12 as an under-explored area rather than a settled evidence base.
• The review's own gap analysis is a limit on what it can recommend: the corpus yielded no concrete teaching-unit examples and no practical daily-training models, so its guidance rests on integration strategies rather than demonstrations.
• Discipline coverage is skewed toward STEM, which means claims about arts, humanities, and creative subjects are extrapolations the review asks future empirical work to test.
• The identified gaps are described rather than quantified: the review reports no pooled effect sizes, and long-term impact on educational outcomes is named as future work rather than an established finding.

Citation

Marzano, D. (2026). Generative Artificial Intelligence (GAI) in Teaching and Learning Processes at the K-12 Level: A Systematic Review. Technology, Knowledge and Learning, 31, 789–829.