Abstract: Currently, artificial intelligence is actively developing and being integrated into all aspects of human life, including science, resulting in both threats and opportunities that require careful consideration. Artificial intelligence can significantly influence the quality and effectiveness of sociological research. This article is dedicated to analyzing and describing the experience of using artificial intelligence, particularly the GPT-4o mini model, in sociological studies for processing and analyzing qualitative data obtained from semi-structured interviews. The author examines the possibilities and limitations of applying GPT-4o mini as a tool for transcribing and analyzing interviews, with special attention given to identifying its analytical capabilities for generating summaries and conclusions. The empirical foundation of the research consists of the results from expert semi-structured interviews focused on marketing ethics conducted among marketing specialists (n=12, 2024). Reflecting on the research experience revealed both positive and negative consequences of utilizing GPT-4o mini for processing and analyzing interviews. The findings indicate that, despite certain limitations in interpretive depth, the GPT-4o mini model can be beneficial for researchers in several aspects, especially regarding the automation of the transcription process. It is essential to note that any interview transcript generated using artificial intelligence must undergo careful verification by the researcher, as the model may encounter difficulties in interpreting human language. Issues such as spelling errors, misattributions of speech, and intonational pauses can lead to a loss of semantic integrity in the text. While assessing the capabilities of GPT-4o mini in data analysis, the model demonstrated mediocre results, typically producing superficial conclusions. In summary, the results of the study can be utilized to optimize the process of processing and analyzing qualitative data in sociological research by incorporating artificial intelligence, particularly the GPT-4o mini model.

Keywords: opportunities of GPT-4o mini, limitations of GPT-4o mini, analytical capabilities of GPT-4o, generation of Conclusions, transcription, GPT-4o mini model, interviews, qualitative Data, sociological research, artificial intelligence