Power, Emotion and AI: A Comparative Media Analysis Across Asymmetric Technological Ecosystems

• In Event: Communication, Information Technologies, and Media Sociology Section Roundtables
  In Refereed Roundtable (60 minute): 15. Media Sociology

Abstract

As artificial intelligence (AI) assumes a central role in geopolitical contestation and in the consolidation of the power of major technology companies, media discourse emerges not merely as a reflection of these transformations but as an active space for narrative construction around the governance, risks and opportunities of AI. However, comparative studies are lacking on how these discourses diverge between central and peripheral nations of the global technological order and on how those differences are structured by geopolitical inequalities, economic dependence and asymmetries in digital governance. This article analyses how news outlets frame the relationship between AI and politics, asking why the media in technologically dominant countries emphasise risks while outlets in dependent economies emphasise development potential. Based on a corpus of 6,829 articles published between 2010 and 2024 in twelve outlets, the study combines topic modelling with sentiment analysis to identify discursive patterns and affective variations. The results reveal three phases: low intensity (2010–2016); politicisation driven by algorithmic interference in elections (2017–2021); and exponential acceleration after the ChatGPT (2022–2024). The dominant frames include the impact of social media (16.03%), the digital economy (12.54%) and the Sino–American geopolitical rivalry (China–United States) (11.41%). The sentiment analysis reveals significant divergences: outlets in central economies produce predominantly negative headlines (41.7%), emphasising surveillance and regulatory risk, while outlets in peripheral contexts present markedly positive coverage (51.5%), highlighting innovation and development. Fear and distrust characterise the framing in central economies; anticipation defines the peripheral coverage. These divergences suggest that media narratives of AI are conditioned by structural asymmetries in the distribution of technological capabilities, functioning as mechanisms of legitimation: of scepticism and regulatory caution where technological power is concentrated, and of developmental aspiration where dependency persists.
Keywords: Artificial intelligence; media framing; digital governance; technological dependency; AI governance; power asymmetries

Author

• Bárbara Maria Farias Mota, Freie Universität Berlin