News & Events


The Use of Generative AI in Education

The Use of Generative AI in Education

Trustee (Vice President)
IIDA Hiroyuki

To Students and Faculty Members,

Given the rapid advancement of artificial intelligence (AI) technology in recent years, there has been an emergence of "generative AI" tools capable of producing texts, images, audio, program codes, etc. As their accuracy and quality continue to improve, concerns in society are growing about their use. User-friendly environments for using generative AI are being established, and generative AI will undoubtedly permeate into our daily lives. Skillful utilization of generative AI holds the potential to partially complement human intellectual activities, leading to significant achievements in both academe and industry.

The mission of JAIST is to foster leaders capable of contributing discoveries and technologies to the world of science and technology through applying superior education and research in an ideal academic environment. With this goal in mind, we need to enhance the quality of learning, achieve profound understanding, and furthermore foster leaders in society and industry through the productive collaboration between humans and AI. To this end, JAIST supports the utilization of generative AI tools in education. In addition, we encourage students to explore novel and practical applications of generative AI on their own.

At the same time, an essential aspect of graduate studying is developing the ability to independently contemplate and act on problem detection and problem-solving methods. Thus, the primary agents of learning are the students themselves. Improper utilization, of AI, such as entrusting the task of tackling research or academic problems to generative AI, would impede the development of the students' capabilities. We strongly encourage students to acquire sufficient knowledge about the strengths, weaknesses and proper role of generative AI, and learn how to best utilize it properly.

As generative AI is currently faced with various technical issues, and several concerns and risks have been identified; it is therefore important that it be used carefully. Users must take responsibility for the content obtained from generative AI.

Please keep the following points in mind when using generative AI.

●Essential Points to Note in Education

[For students]
・Concerning the use of generative AI in class, follow the instructions given by the lecturers.
・In cases where you employ generative AI for creating a report or working on an assignment, explicitly state the fact that you have used it.
[For faculty]
・Provide explicit instructions and explanations to students about how to utilize generative AI in the lecture in charge.
・Verify in advance how generative AI answers questions in assignments and reports, and design questions so that generative AI cannot easily produce plausible answers.
・To ensure fairness among students, take into consideration the use of generative AI when evaluating and grading students' answers.
[For students and faculty]
・Employing software that incorporates generative AI may lead to inadvertent utilization of generative AI. Pay attention to the software's environment settings.
・Before using generative AI, it is highly recommended to apply for opt-out (prohibition of using input data for AI learning) to avoid unintentional input of personal or confidential information. It is also possible to disable the saving of conversation history to prevent others from using your input data for AI learning. Adjust the settings appropriately.

●Points to Note in Data Input [For students and faculty]

・Do not input highly confidential information such as the following, because input data are disclosed to a third party (the generative AI company).
- Personally identifiable information
- Information subject to non-disclosure agreement (NDA) in collaborative research with companies
- Confidential information of JAIST and laboratories

●Points to Note in the Use of Generative AI Output [For students and faculty]

・A large language model, one of the core technologies in generative AI, produces plausible texts by repeatedly predicting the most likely word after a given word sequence. However, the generated content may potentially include errors and falsehoods.
- Do not accept the output as it is. Always maintain a critical attitude, confirm its accuracy, and make necessary modifications when using it.
- The output may include false personal information. Note the possibility of leading to defamation by using it.
・Using output without any changes in parts of reports or other submissions can be considered plagiarism, and thus an act of academic misconduct. Pay careful attention.
・Using output which is the same as, or similar to, existing copyrighted material may potentially constitute copyright infringement. When utilizing AI output, compare its similarity with existing copyrighted source material.