Use of AI in studies

JYU Open University allows the use of AI in support of studies according to specified guidelines.  Policies are updated as more experience with AI in education is gained.

Various language models and applications based on artificial intelligence have become part of studying. GPT (Generative Pre-trained Transformer) and other similar AI-based text editors or text generators, known as large language models, are interactive AI applications that produce text based on user input. You can prompt the AI application to answer questions and, for example, ask it to rephrase a text in order to improve it.

The language model works on the basis of probabilities: it uses an enormous amount of data to predict which word will come after the previous one. The language model itself does not think, but instead it uses the data to generate text, which can be correct, but also distorted or incorrect.

For this reason, the language model cannot be used as a source of information and the content it generates cannot be blindly trusted. It is always necessary for the user to critically evaluate the material obtained from the language model; the user is responsible for the accuracy of the material.

The JYUOpen teaching development committee has outlined the use of AI in learning as follows:

  1. You are allowed to use AI to support your studies. You can use the language model, for example, to come up with new ideas or create basic texts or drafts, structure ideas, understand concepts, create an overall picture or to help with grammar and editing. In other words, AI can produce raw material and groundwork that you need to process further yourself.
  2. AI is not a scientific source. Each course has a defined bibliography, which will be used as the main source of information. AI can support studying as outlined in the first item, but as such it is not suitable as a source, especially for theses. In keeping with good scientific writing practice, you should always refer to the original source.
  3. All study credits are subject to transparency requirements. To ensure a fair assessment, you must clearly show what constitutes your own thinking and where you have used the ideas obtained from the language model in your coursework. If you have used text formulated by an AI in any part of your work, you must disclose it appropriately.
  4. You are always responsible for the content of the text you submit. If you use a text produced by a language model without editing it, that is considered plagiarism. In cases of plagiarism, the coursework may be marked as failed.

Guidelines

You must disclose the use of AI in all written work. Written work includes assignments, exam answers or any piece of writing that may be produced independently or, for example, as part of an online course (e.g., learning diary, report, essay, online discussion).

Disclose any use of AI in the introduction or at the beginning of the written work. If the description is long, you can also provide it in a separate annex. The description must include the following:

  • The AI-based models or applications you have used (e.g., Microsoft Copilot, ChatGPT, Elephas, GhostWriter, Bing, Rytr.me, Lex.page, Jasper, Notion AI)
  • Why you have used AI (see item 1 of the policy)
  • The prompts you have given to the AI application (search terms and questions, tasks given to the AI application and conversations with it)
  • The date on which you generated the data using AI

The use of artificial intelligence must also be indicated in the bibliography and mentioned in the written work itself.

A teacher may, for a justified reason, prohibit the use of AI in an individual assignment or course. In such a case, the teacher should explain the exceptions and the reasons for them clearly to the students.

These guidelines and policies will be updated as more experience is gained in the use of AI.