AI in K-12 Classrooms: A Structured Governance Framework for Cognitive Preservation

Seung Ho Jeon, Hrishikesh Desai, H. Steve Leslie
Proceedings of the Impactful and Responsible AI Systems for Education Workshop, PMLR 339:194-202, 2026.

Abstract

The rapid integration of Artificial Intelligence (AI) into K-12 education presents both significant instructional opportunities and emerging governance challenges. While existing research has largely focused on the technical capabilities or theoretical implications of AI in education, limited empirical work captures how educators themselves perceive its use in real classroom settings, particularly at the district and state levels. This study addresses that gap by examining educator perspectives on AI adoption across nine school districts in Arkansas, in collaboration with the Arkansas Department of Higher Education. Using an IRB-approved survey instrument, data were collected from 46 educators across five domains: instructional and administrative benefits, classroom risks and downsides, trust and ethical considerations, governance and oversight structures, and professional support and training needs. Findings indicate strong recognition of AI’s efficiency and instructional value, particularly in lesson planning, grading, and content generation. However, these benefits are accompanied by substantial concerns, including the loss of critical thinking, risks to academic integrity, misinformation, and data privacy risks. Notably, results reveal a pronounced gap between individual-level AI usage and institutional-level trust, with educators expressing strong demand for structured oversight, standardized tools, and formal training. Building on these findings, this study proposes a structured governance framework for AI integration in K-12 education that links technological capability, risk emergence, institutional trust, and policy design to the preservation of student cognitive autonomy. The framework emphasizes that the central challenge of AI adoption is not technological readiness, but the absence of coordinated governance mechanisms that align AI use with educational objectives. By providing localized, policy-relevant evidence and a conceptual model for responsible integration, this study contributes to ongoing discussions on AI governance, educational standardization, and the future of teaching and learning in the AI era.

Cite this Paper

BibTeX
@InProceedings{pmlr-v339-jeon26a, title = {AI in K-12 Classrooms: A Structured Governance Framework for Cognitive Preservation}, author = {Jeon, Seung Ho and Desai, Hrishikesh and Leslie, H. Steve}, booktitle = {Proceedings of the Impactful and Responsible AI Systems for Education Workshop}, pages = {194--202}, year = {2026}, editor = {Basu Mallick, Debshila and Woodhead, Simon and Wang, Zichao and Ananda, Muktha and Burstein, Jill and Murphy, April}, volume = {339}, series = {Proceedings of Machine Learning Research}, month = {28 Jun}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v339/main/assets/jeon26a/jeon26a.pdf}, url = {https://proceedings.mlr.press/v339/jeon26a.html}, abstract = {The rapid integration of Artificial Intelligence (AI) into K-12 education presents both significant instructional opportunities and emerging governance challenges. While existing research has largely focused on the technical capabilities or theoretical implications of AI in education, limited empirical work captures how educators themselves perceive its use in real classroom settings, particularly at the district and state levels. This study addresses that gap by examining educator perspectives on AI adoption across nine school districts in Arkansas, in collaboration with the Arkansas Department of Higher Education. Using an IRB-approved survey instrument, data were collected from 46 educators across five domains: instructional and administrative benefits, classroom risks and downsides, trust and ethical considerations, governance and oversight structures, and professional support and training needs. Findings indicate strong recognition of AI’s efficiency and instructional value, particularly in lesson planning, grading, and content generation. However, these benefits are accompanied by substantial concerns, including the loss of critical thinking, risks to academic integrity, misinformation, and data privacy risks. Notably, results reveal a pronounced gap between individual-level AI usage and institutional-level trust, with educators expressing strong demand for structured oversight, standardized tools, and formal training. Building on these findings, this study proposes a structured governance framework for AI integration in K-12 education that links technological capability, risk emergence, institutional trust, and policy design to the preservation of student cognitive autonomy. The framework emphasizes that the central challenge of AI adoption is not technological readiness, but the absence of coordinated governance mechanisms that align AI use with educational objectives. By providing localized, policy-relevant evidence and a conceptual model for responsible integration, this study contributes to ongoing discussions on AI governance, educational standardization, and the future of teaching and learning in the AI era.} }
Endnote
%0 Conference Paper %T AI in K-12 Classrooms: A Structured Governance Framework for Cognitive Preservation %A Seung Ho Jeon %A Hrishikesh Desai %A H. Steve Leslie %B Proceedings of the Impactful and Responsible AI Systems for Education Workshop %C Proceedings of Machine Learning Research %D 2026 %E Debshila Basu Mallick %E Simon Woodhead %E Zichao Wang %E Muktha Ananda %E Jill Burstein %E April Murphy %F pmlr-v339-jeon26a %I PMLR %P 194--202 %U https://proceedings.mlr.press/v339/jeon26a.html %V 339 %X The rapid integration of Artificial Intelligence (AI) into K-12 education presents both significant instructional opportunities and emerging governance challenges. While existing research has largely focused on the technical capabilities or theoretical implications of AI in education, limited empirical work captures how educators themselves perceive its use in real classroom settings, particularly at the district and state levels. This study addresses that gap by examining educator perspectives on AI adoption across nine school districts in Arkansas, in collaboration with the Arkansas Department of Higher Education. Using an IRB-approved survey instrument, data were collected from 46 educators across five domains: instructional and administrative benefits, classroom risks and downsides, trust and ethical considerations, governance and oversight structures, and professional support and training needs. Findings indicate strong recognition of AI’s efficiency and instructional value, particularly in lesson planning, grading, and content generation. However, these benefits are accompanied by substantial concerns, including the loss of critical thinking, risks to academic integrity, misinformation, and data privacy risks. Notably, results reveal a pronounced gap between individual-level AI usage and institutional-level trust, with educators expressing strong demand for structured oversight, standardized tools, and formal training. Building on these findings, this study proposes a structured governance framework for AI integration in K-12 education that links technological capability, risk emergence, institutional trust, and policy design to the preservation of student cognitive autonomy. The framework emphasizes that the central challenge of AI adoption is not technological readiness, but the absence of coordinated governance mechanisms that align AI use with educational objectives. By providing localized, policy-relevant evidence and a conceptual model for responsible integration, this study contributes to ongoing discussions on AI governance, educational standardization, and the future of teaching and learning in the AI era.
APA
Jeon, S.H., Desai, H. & Leslie, H.S.. (2026). AI in K-12 Classrooms: A Structured Governance Framework for Cognitive Preservation. Proceedings of the Impactful and Responsible AI Systems for Education Workshop, in Proceedings of Machine Learning Research 339:194-202 Available from https://proceedings.mlr.press/v339/jeon26a.html.

Related Material