The Shaping of the Image of Artificial Intelligence as a New Public Challenge in the Russian Mass Media

Neural networks and artificial intelligence (AI) technologies have become part of everyday news discourse due to the rapid growth of popularity and availability of these technologies for mass use. Due to the trend towards the introduction of generative technologies in any sphere of production, the interest in this technology in the public space is also growing. One of the sources of information about innovations is the media. As long as the technology is not in mainstream use, people's opinions are shaped, in part, by the media and opinion leaders. The ways that journalists use to describe AI technologies can highlight the potential risks or, conversely, the potential benefits of AI technologies in the future. Thus, the aim of the study is to determine how the news agenda about artificial intelligence is framed in Russian-language media. Qualitative content analysis was chosen to achieve the objective. This method allows analyzing large volumes of text published in the media by dividing the text into units of analysis and following a theoretical categorical framework in order to search for meanings in the texts themselves and in the context of the publication. A news article published by a journalist or a group of journalists in a media outlet (online edition, newspaper, magazine or news agency) of unregulated length was taken as the unit of analysis; the text should also contain a reasoned position or opinion of a public personality with access to decision-making in politics or business regarding AI technology. Based on the trend line of the number of news articles published over the last two years, we determined the chronological framework of the empirical news base, which is located in the time period from November 1, 2023 to March 31, 2024. During this time period, the information field was most active in publishing news articles mentioning artificial intelligence in the headlines and texts of articles. After excluding from the database duplicates of publications in mass media (online publications, newspapers, magazines and news agencies) we received 15,997 unique news articles, which made up the general data set of the study. From the general data set, a sample of 210 news stories was selected for analysis by stratified-random method with specified quotas according to the percentage distribution by media level and category in the sample population. The text coding process consisted of two stages: pilot coding with data-driven refinement of the coding sheet and main coding according to the fixed coding sheet. Both manifest (according to L. Newman) and latent coding were used in the coding process. According to the results of the analysis, the initial hypothesis that artificial intelligence is presented in the media as a public challenge and threat was not confirmed. Frames of neutral and positive tone rather prevail in the media, as two thirds of the articles included in the sample are neutral or moderately optimistic in their judgments. By structure, the analyzed articles contain assumptions about benefits, such as Automation and reduced labor costs and increased labor efficiency, and risks, such as Loss of control over technology, Mass job losses and unemployment; Data leaks and cyberattacks; Lack of legal regulation and discrimination.