Recommendation for Creative Activities Area

In the field of creative activity at the BUT, it is necessary to apply the basic principles of using generative AI tools issued by the university.

Open access, scalability, transparency and reproducibility

The use of generative AI tools in the activities of academic and research staff is encouraged at the BUT.

Employees are entitled to use generative AI tools in the context of producing a scientific result or artistic output, but must indicate which tools have been used and to what extent.

Recommendations for how to indicate the use of generative AI tools in publications are constantly evolving, so keep an eye on the current situation. Currently, the guide "How do I cite generative AI in MLA style? [1]"[2] , published by the Modern Language Association (MLA), which summarises how to cite generative AI in MLA style. For other citation styles, the ChatGPT article ChatGPT Citations offers a summary of citation styles. [3]

At the same time, it is the responsibility of the staff member who used the tools to check the accuracy of third-party citations included in the output produced by the generative AI tools. It is recommended to record and document the use of the generative AI tools, including the parameters and inputs that were used and the outputs that were received, while indicating the version of the tool that was currently used. This allows for transparency and in some cases reproducibility of results, and makes it easier for other researchers to verify and examine the analyses performed. [4]

Control of terms of use and copyright

It is necessary for the employee to verify the (contractual) terms of use of a specific generative AI tool, which are always formulated by the provider of the tool.

These terms may also stipulate the authorship of the outputs of the tool, thus verifying that the employee does not infringe the copyright of the provider.

Critical preview of outputs

It is recommended to look critically at the information and outputs that generative AI tools provide. Information should not be taken without verification. In the case of ChatGPT, for example, it is advisable to ask for and verify the URLs on which the tool relied to generate the text.

Caution when inserting data

When inserting data into generative AI tools, it is important to make sure that the employee has permission to use the data they are sharing in this way. Special care should be taken when uploading personal and other sensitive data.

[1] How do I cite generative AI in MLA style?. 2023. [cit. 2023-6-27]. Dostupný z WWW: [https://style.mla.org/citing-generative-ai/?utm_campaign=sourcemar23&utm_medium=email&utm_source=mlaoutreach].

[2] MAŠEK, Jaroslav. Kdy a jak citovat AI podle MLA. <em>Metodický portál: Spomocník </em>[online]. 07. 04. 2023, [cit. 2023-06-27]. Dostupný z WWW: <https://spomocnik.rvp.cz/clanek/23452/KDY-A-JAK-CITOVAT-AI-PODLE-MLA.html>. ISSN 1802-4785.

[3] ChatGPT Citations | Formats & Examples. Scribbr [online]. Amsterdam: Scribbr, 2023 [cit. 2023-06-27]. Dostupné z: https://www.scribbr.com/ai-tools/chatgpt-citations/

[4] Jaká jsou pravidla pro odpovědné používání nástroje ChatGPT a dalších nástrojů umělé inteligence pro vědce a akademiky”. ChatGPT, 24 May version, OpenAI, 26 June. 2023, chat.openai.com/chat.