For students

AI can be a useful tool, for example, for

• explaining concepts,

• obtaining feedback,

• creating self-assessment questions,

• brainstorming,

• assisting with programming,

• editing content, language revision and translation,

• practising critical thinking (e.g. evaluating AI-generated responses).
  

What matters is the learning process: the ability to write good prompts, critically assess the quality of the output, and engage in meaningful dialogue with AI.

If you submit work created wholly or partly using AI under your own name without stating and describing the use of AI, this is considered academic fraud.

If you violate the AI use restrictions described in the course information but have transparently acknowledged the use of AI, it may not be considered fraud, but it may still mean that you have not met the assessment criteria, and the teaching staff member may not accept the work or may lower the grade.

Personal responsibility is an integral part of academic freedom.

When using AI, consider data protection and confidentiality requirements, ensure data security, and verify the AI tool's suitability for the task. Therefore

• avoid entering personal data or other sensitive information into AI applications,

• do not copy sensitive documents or their detailed contents into prompts,

• read the data protection terms of the application,

• follow the university’s recommendations when selecting secure tools.
  

If processing sensitive information is necessary, prefer AI applications that can be used with a University of Tartu account and have been approved by the IT Office. Where possible, use anonymised data.

If you use AI, clearly describe

• what the AI helped to produce (e.g. text sections, tables, figures, ideas),

• which tool you used (name, provider/creator, version, etc.),

• when you used it (date).

AI must not be credited as an author or co-author, as only a human can be an author.

Tools that merely edit or format your own text (e.g. translation and editing tools or reference managers) do not necessarily need to be cited if they do not create new content. However, different disciplines and teaching staff members may have different approaches.

Using AI as a support tool at different stages of thesis writing is not prohibited, but presenting AI-generated output as your own ideas is academic fraud and contrary to the code of conduct for research integrity.

You must be able to justify the substantive choices made in your thesis. You must also describe the use of AI (e.g. in the methodology chapter) with sufficient precision so that it is clear to what extent and for what purpose it was used.

AI may be helpful in thesis writing, for example, for

• ideation,

• refining structure,

• finding synonyms,

• identifying relevant sources,

• translation,

• understanding complex concepts,

• obtaining feedback,

• identifying programming issues.
  

AI must not be used to generate large sections of a thesis (e.g. a full chapter or summary) or substantive lines of argument. The University of Tartu considers this academic fraud and a violation of the code of conduct for research integrity.