Series Index
Part 1 | When AI Enters the Classroom, What Choice Will Korean Education Make?
Part 2 | AI Has Broken Korea’s School Assessments
Part 3 | Why Korea Needs a New Framework Act on Education in the AI Era
By Jung Yeon-Tae,
President of the Korea Information Service of the Construction Industry and Essayist.
Artificial intelligence is already in the middle of the classroom. Students now ask AI instead of textbooks, and teachers are asked less to explain and more to judge. But while change accelerates in schools, the legal foundation of education still sits on older assumptions. That foundation is the Framework Act on Education.
A common objection is rising: “If there are already AI and digital-education bills pending, why do we need a new Framework Act on Education?” The question is reasonable. The National Assembly is considering multiple proposals, including amendments to the Framework Act, a special act for digital-based public education innovation, and revisions to the Distance Education Act. The core issue, however, is not the number of bills, but what they define—and what they leave undefined.
Most pending proposals function as implementation laws. The current amendment to the Framework Act adds AI-related provisions so national and local governments can promote AI capability and ethics education for students and teachers. It is necessary, but it does not redefine the purpose of education or rebuild the rights-and-responsibilities architecture. In effect, it authorizes what to do, without answering what education is.
The Digital Public Education Innovation Special Act is broader. It addresses digital textbooks, learning-data processing and protection, national integrated systems, teacher training, and infrastructure expansion. The Distance Education Act revision also aims to elevate digital and media literacy into a national plan. Yet these bills share a design logic centered on operations: they answer “how to run AI education,” not “what education should become in the AI era.”
A framework act is meant to set direction. It declares the purpose of education, learner rights, and state duties, functioning as a charter rather than a program manual. If this charter is out of date, implementation laws accumulate without restoring coherence. In the AI era, the central question is unavoidable: if AI explains knowledge and solves problems, what should humans learn?
The answer cannot be limited to technical adaptation. Education must shift from memorization and reproduction to judgment and responsibility—deciding whether to accept or reject machine-provided answers, explaining why, and bearing consequences. That is a value choice, not merely a curriculum tweak. It belongs in the Framework Act’s core declarations.
The legal vacuum becomes clearer when learner rights are examined. Some bills address learning-data protection and related data rights, which is progress. But AI-era learner rights must also include the right to demand explanations for AI-recommended learning paths or AI-influenced evaluation outcomes, the right not to be compelled to use AI, and the right to request human teacher intervention in high-stakes decisions. Remedies for harm caused by algorithmic bias or system errors also remain largely undefined.
Responsibility is the other missing pillar. If AI misjudgment distorts learning opportunities or automated evaluation produces major disadvantages that affect a student’s trajectory, who is accountable? Is it the teacher, the school, the system operator, the platform provider, or the state? Without clear responsibility lines, schools become risk-averse, teachers hesitate, and trust in outcomes erodes. AI education then leaves behind uncertainty rather than improvement.
This is why more laws can mean a stronger need for a higher law. As amendments to the Framework Act, special acts, and distance-education revisions move on separate tracks, the field struggles to identify the controlling standard. Students and parents may not know where to claim rights, and teachers may not know where duty ends. Fragmentation increases confusion, and overlapping rules multiply compliance burdens.
Comprehensive legislation is not duplication for its own sake. It can be the act of organizing duplication, placing operational rules under a coherent hierarchy of principles. The point is not to dismiss pending bills, but to recognize their value and build an integrated charter above them. That requires more than partial edits; it requires a near-full rewrite of the Framework Act on Education for the AI era.
At minimum, a new framework act should do four things. It should redefine education’s purpose around human dignity, judgment, responsibility, and meaning-making capacity. It should establish a learner-rights system that includes explanation rights, refusal rights, and meaningful human oversight, not only data rights. It should redefine teachers as learning designers and ethical stewards of AI use while clearly delimiting their responsibility so they do not absorb systemic risk alone. And it should define the state as the guarantor of publicness, the corrector of inequality, and the ultimately accountable entity when the system causes harm.
Education is not meant to trail technology. It is designed with the future in mind, yet the current Framework Act remains anchored in the past. Leaving it unchanged in the AI era is akin to regulating autonomous vehicles with laws written for the steam engine. The pending bills are necessary, but they are not sufficient to answer the foundational question.
The country must now ask—and answer—its most basic educational question: in the AI era, what kind of human being are we trying to cultivate? The state’s official answer should be a new Framework Act on Education. Without that charter, implementation statutes will keep piling up while the system’s direction remains unresolved.
Author Bio
Jung Yeon-Tae is a public policy specialist who has studied policymaking and legislation at the intersection of law and institutions, technology and education. Drawing on experience with legislative processes at the National Assembly and public-sector administration, he has consistently proposed redesigns of Korea’s national education framework and legal system for the AI era. He currently leads a public agency under the Ministry of Land, Infrastructure and Transport and continues to write and publish under the pen name “Manripungchwi,” focusing on civilizational transition, education, and the relationship between technology and the human condition.
Coming next
Part 4 | Teachers, Face, and the Trap of Institutions—Why Educators Fall Silent in Front of AI
Korean version available here.
