Data and code for a bibliometric and topic-modelling study of the economic sustainability of news media in the age of AI

This dataset supports a bibliometric and topic-modelling study on the economic sustainability of news media. It includes the raw Scopus (n = 705) and Web of Science (n = 455) exports in BibTeX format, retrieved through a keyword search and a targeted search of three core media-economics and media-management journals (Journal of Media Economics, International Journal on Media Management and Journal of Media Business Studies); the screening log with the reason for every decision; the screening audit decisions; the cleaned corpus of 518 publications (1987–2026); the analysis results; and the Python code with requirements.txt (Python 3.13.5). Run the scripts in numerical order (01 → 04). Script 02 generates the 16 figures; script 03 performs additional robustness checks; script 04 runs the sensitivity analysis excluding records that address news together with other media sectors. Data were frozen on 30 October 2026.