Set of questions used for expert email interviews on the use of GenAI in terminology work conducted in 2025 by Tanja Wissik within the project "Study on Terminology Workflows in the 21st Century"

Set of questions used for expert email interviews on the use of GenAI in terminology work conducted in 2025 by Tanja Wissik within the project “Study on Terminology Workflows in the 21st Century” within the second part of the study. Tanja Wissik Study on Terminology Workflows in the 21st Century 2 Question Set A (for experts with greater involvement with the content aspect of terminology workflows) 1. Does your institution/unit have a policy/guidelines regarding the use of applications based on (generative) AI or LLMs? Or is such a document currently being developed? 2. Have you tried to use ChatGPT, Gemini or other similar generative AI applications for some of the steps in terminology work? What was your experience with them? How would you rate the quality of the results for the languages you are working with? 3. Do AI and large language models (LLMs) play a role in your daily terminology work? Do you use tools based on AI or LLMs or that have integrated AI modules for terminology management, data cleaning, term extraction, term translation, definition writing or other tasks? a) Please name and describe the tool(s) and the task you use them for. b) How was your experience with the tool(s) (pros and cons)? c) How do they integrate into your established workflow and with the other tools you use? d) Did you experience differences in quality for different languages? Question Set B (for IT-experts involved in terminology workflows) 1. Does your institution/unit have a policy/guidelines regarding the use of applications based on (generative) AI or LLMs? Or is such a document currently being developed? 2. Do (generative) AI and large language models (LLMs) play a role in your daily work in developing terminology and linguistic tools? Do you use tools based on (generative) AI or LLMs for developing terminology management systems or other linguistic tools? Do you integrate (generative) AI or LLMs based applications into the tools you are developing. If so, for which task you integrate AI? If so, do you use already existing LLMs out of the box, do you finetune existing LLMs or do you train your own? 3. Do you get requests from users to integrate (generative) AI or LLM based applications into the systems you develop? If yes, for which tasks?