Japan's education ministry on Tuesday unveiled guidelines allowing the limited use of generative artificial intelligence, such as ChatGPT, in elementary, junior high and high schools.
The guidelines released by the Ministry of Education, Culture, Sports, Science and Technology call for particular caution regarding use by elementary school students, while stating that passing off AI-assisted school work as one's own will be deemed cheating.
While recognizing the importance of students gaining a firm grasp of AI and its uses, the guidelines also took into account concerns the technology may negatively impact students' critical thinking and other skills.
The guidelines are provisional and will initially only allow for limited use.
The education ministry will select by fall a number of junior high and high schools that will trial AI use and plans to revise the guidelines based on the results.
It is expected AI will improve educational outcomes, but utilizing the technology carries risks of personal data leaks and copyright infringements and could stifle students' creativity and motivation to learn, the guidelines said.
They also state it is necessary to educate children on ethical issues associated with AI, given that it can be used anywhere.
The guidelines outline examples of inappropriate AI use, including students passing off AI-generated work as their own or using the technology during exams.
It is also important that students do not put personal information into AI tools and observe copyright rules, the guidelines said.
The ministry suggests AI should be used to gain additional viewpoints for class discussions. Students and teachers should also look at the limitations of the technology by assessing the nature of any erroneous information generated.
But despite the guidelines not specifically mentioning it, education ministry officials have suggested elementary school students younger than 13 can use AI to a certain degree under the guidance of teachers.
ChatGPT, developed by U.S. tech firm OpenAI, and other generative AI programs are trained using massive amounts of data from the internet and can process and simulate human-like conversations with users or create images based on users' instructions.