Visitors observe AI-related education tool at an exhibition held at Coex convention center in Gangnam District, southern Seoul, on Jan. 21.YONHAP
More than half of Korean schoolteachers now use AI at least weekly, but only about one in six are familiar with the laws and guidelines governing it, leaving individual teachers to manage hallucinations, privacy risks and ethical questions largely on their own, according to a Korean Educational Development Institute report released Monday.
The survey of 271 elementary, middle and high school teachers nationwide found that 54.6 percent use AI for work at least weekly. Another 17.3 percent said they use it several times a day, while only 3 percent never do.
Teachers overwhelmingly favored general-purpose generative AI platforms such as ChatGPT, Claude and Gemini, which 94.5 percent of respondents used.
Another 80.4 percent said they used AI-powered content creation tools, while 39.5 percent reported using educational courseware, learning management systems and learning support software.
Teachers most commonly used AI to generate lesson ideas and plan classroom activities, as cited by 39.2 percent of respondents. Another 36.2 percent used it to create or revise teaching materials, worksheets and presentations, while 28.7 percent relied on it to draft official documents, reports and meeting minutes.
Its use was far less common for activities that involved direct interaction between students and AI or the evaluation of student work.
The researchers attributed the pattern to teachers' preference for using AI in tasks where they could directly verify and refine its output.
Visitors attend a training session on AI-powered digital educational materials at an exhibition held at Coex in Gangnam District, southern Seoul, on Aug. 12.NEWS1
The risks behind AI adoption
AI hallucinations, in which the technology generates false information and presents it as fact, were the most frequently reported problem among teachers. Some 64.2 percent of teachers reported encountering AI hallucinations or a lack of transparency in how AI generates responses, the highest share among all problems surveyed.
Beyond accuracy and transparency, 42.8 percent of teachers feared that AI could undermine student learning, while 40.6 percent pointed to risks associated with the handling of personal information, academic data and sensitive information.
Ethical challenges, from copyright violations to improper attribution and plagiarism, affected 35.4 percent of teachers. Nearly 30 percent identified disparities in students' access to AI and digital proficiency as a problem.
AI has become a routine part of teachers' work, but their knowledge of the rules governing it has failed to keep pace.
Only 17.3 percent of teachers said they were familiar with the laws, policies and guidelines governing AI use, while 45.4 percent said they were unaware of them, roughly 2.6 times as many.
Researchers cautioned that responsibility for addressing AI-related problems in schools falls disproportionately on individual teachers and their professional judgment. Greater reliance on the technology could add to their workloads as they must verify its output and respond to potential problems.
To address these challenges, the researchers recommended clearly defining who should oversee AI use and where responsibility lies. They also called for a system to document and report problems when they arise and take corrective action.
The researchers also called for a dedicated body to continuously monitor AI use in schools and adjust relevant policies. They proposed establishing a central authority to coordinate and oversee AI education policy.
This article was originally written in Korean and translated by a bilingual reporter with the help of generative AI tools. It was then edited by a native English-speaking editor. All AI-assisted translations are reviewed and refined by our newsroom.
