Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
South Korea’s national AI-textbook experiment unraveled as a policy and procurement project: after launching in March 2025, the products were reclassified from official textbooks to optional educational materials in August. A later government audit criticized the preparation, including the absence of a proper pilot and inadequate field-suitability review. That is a serious failure—but it does not establish that every product harmed learning or that AI has no place in schools.
What South Korea set out to do
Under former President Yoon Suk Yeol’s administration, South Korea proposed using AI-enabled digital learning platforms to personalize instruction. The systems were intended to analyze learning patterns, identify gaps, recommend exercises and give teachers information to help differentiate lessons. The aim was to bring those tools into ordinary classrooms at national scale.
This was not simply a plan to put ChatGPT into printed textbooks. The products combined curriculum content with digital learning functions, adaptive activities and data features. The ministry described them as AI digital textbooks, but products could differ, and the available evidence does not establish that all generated explanations or textbook prose. Some personalization functions may also have relied on conventional software rather than generative AI.
The first rollout targeted elementary grades 3 and 4, middle-school grade 1 and high-school grade 1, in English, mathematics and information/coding. The government’s original expansion ambitions were broader, but in November 2024 it narrowed the initial 2025 launch. The Ministry of Education’s rollout description sets out the grades and subjects; the government’s revised roadmap describes the narrower launch.
#1 Best Overall
The ministry established a regulatory basis for AI digital textbooks in 2023 and created a review process. After 146 products from 21 applicants were submitted, it approved 76 titles from 12 publishers for the 2025 school year. Approval did not mean every school adopted them, or that adoption proved their effectiveness. The ministry’s explanation of the legal framework and its approval announcement document those steps.
A national launch, but uneven use
The products entered classrooms with the start of the 2025 academic year. OECD reporting puts initial adoption at roughly one-third of schools using at least one AI textbook. The rate varied sharply by region: reported figures included 98% in Daegu and 8% in Sejong. Those figures describe whether schools adopted at least one product, not how often teachers used it, whether students used it throughout lessons, or whether the product improved learning. The OECD’s account is useful precisely because it distinguishes rollout from universal classroom use.
The distinction matters. A list of approved titles measures what was cleared for use; adoption measures school selection; regular use and learning outcomes require different evidence. The audit report examined selection, use frequency, user experience and reasons for non-use or discontinuation, but the available figures do not justify treating the national rollout as either universal success or universal classroom chaos.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Why the rollout became contentious
Opposition did not represent one uniform position. Some educators and supporters saw potential in adaptive practice and teacher dashboards. Critics—including teacher organizations and civic groups—objected to compulsory or rapid adoption, questioned whether benefits had been demonstrated, and raised concerns about screen exposure, privacy, teacher workload, legal authority and cost. A 2024 academic analysis also identified concerns about training, educational inequality and the possibility that digital initiatives could add to private-education costs (academic analysis).
These objections should not be collapsed into a claim that South Korean teachers rejected AI as a concept. Opposition to a national mandate is different from opposition to optional trials; concerns about a particular product or timetable are different from opposition to educational technology altogether. A system can be promising in principle while still being poorly timed, burdensome or inadequately tested in practice.
The implementation also had practical questions that matter in any classroom technology project: whether school networks and devices were dependable, whether teachers had enough time and training, how student data was handled, and what happened when a platform failed or gave a poor explanation. These are important tests, but they should not be presented as confirmed nationwide incidents without product- or school-level evidence.
Rank #3
The legal reversal changed the business model
In August 2025, the National Assembly changed AI digital textbooks’ status from official textbooks to optional educational materials. That reversed the central premise of the rollout. Schools still had to meet curriculum requirements using conventional textbooks; the AI versions became supplementary choices rather than required replacements. The Ministry of Education said it would halt the review process for products intended for the 2026 school year. Yonhap’s report on the legislative vote and its report on the paused review describe the change; the OECD summary explains its practical implications.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11That status change affected more than classroom choice. The initial policy had offered publishers a government-backed route to broad school procurement. Once the materials became optional, adoption and subscription revenue became less predictable, while schools faced questions about who would pay. The reversal therefore created commercial risk for publishers and ed-tech companies that had invested in a market shaped around official status and expected public purchasing.
That is not the same as proving a particular amount of public waste or private loss. A proper accounting would distinguish development grants, teacher training, devices and network upgrades, review costs, school subscriptions, unused licences and publisher investment. The audit record includes budget and subscription-cost analysis, but the evidence cited here does not establish a single verified total for waste. Large figures circulating without such accounting should not be repeated as fact.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.The audit points to a process failure
A December 2025 government audit made the rollout’s preparation—not simply its partisan reception—the central issue. Its reported findings included that a proper pilot operation was omitted, field-suitability review was insufficient, technical standards had not been adequately prepared before the review announcement, and provincial education authorities had not been sufficiently consulted about subscription costs. The audit also examined school selection, usage and reasons for non-use or discontinuation. The National Assembly Library record identifies the report and its contents.
“No proper pilot” should not be stretched into “no one ever saw or demonstrated the products.” It is a finding about the absence of a proper pilot operation, not proof that there were no demonstrations, training sessions or limited exposure. But the distinction does not blunt the finding: for a nationwide change to classroom practice, demonstrations are not a substitute for pilots that test ordinary school conditions, teacher usability and student experience before scale.
Free tools Windows power users keep installed
One-click scans. No signup required.
The underlying risks are familiar in public technology procurement. A robust rollout would ordinarily test accessibility and usability, cybersecurity and privacy protections, network resilience, teacher workload and the process for correcting errors. It would also check whether benefits justify ongoing subscription and maintenance costs, and whether schools can return to conventional materials without losing continuity. The audit findings indicate shortcomings in piloting, field review, technical preparation and cost consultation; they do not by themselves establish that every other possible failure occurred.
Best Value
What “disaster” does—and does not—mean
The record supports calling this a policy and commercial failure. The government’s official-textbook model was reversed within months of launch; adoption was far from universal and varied widely by region; and the audit criticized key elements of preparation. The commercial model was weakened when expected public-backed procurement became uncertain.
It is a different claim to say that AI textbooks were shown to reduce learning or that students were broadly harmed. The rollout was curtailed before it produced a mature, widely accepted independent evidence base on learning outcomes. That leaves a serious unanswered question, not a verdict that all the products were pedagogically ineffective. Likewise, uneven adoption raises an equity concern but does not, on its own, prove whether the tools reduced or widened educational gaps.
The broader lesson is not that South Korea abandoned AI in education, or that digital learning cannot help students. It is that a technology’s promise does not settle whether it should be made a national procurement priority. Personalization requires student data; cloud platforms can offer useful features but depend on reliable networks; dashboards may inform teachers but can also add monitoring work. National standards can promote consistency, while a mandate can magnify flaws before they are understood.
Recommended Free Tools
What other school systems should take from it
- Pilot before scaling. Test products across ordinary schools, regions and student needs, not only in showcase settings.
- Measure outcomes independently. Track learning and retention, teacher workload, accessibility, usage and costs—not just devices deployed or accounts activated.
- Make procurement reversible. Schools need a clear way to stop or change products without disrupting curriculum obligations.
- Set legal and financial responsibilities early. Define who pays, who controls student data, who corrects errors and who is accountable for long-term support.
- Budget for the whole service. Training, connectivity, maintenance, technical support and subscriptions matter as much as launch costs.
- Protect choice and equity. A system should not assume that every school has the same infrastructure or capacity to implement it.
South Korea’s episode is best understood as a warning about turning a technology ambition into a nationwide mandate before evidence, school readiness, legal authority and sustainable financing are settled. The experiment’s reversal is clear; the educational verdict on AI-assisted learning remains more open.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
