Imagining Regulation of AI: Comparing Czech and Portuguese Media Representations with Cads

Artificial Intelligence (AI) has become an inevitable part of the visions of the digital future. Thus, the possibility of AI regulation is one of the most crucial questions in shaping it. The development of technologies like AI depends on media-shaped public perception (Chuan et al., 2019). This research focuses on media debates about AI regulation from an interdisciplinary comparative perspective. It analyses imaginaries in the Czech and Portuguese online mainstream media, using Corpus Approaches to Discourse Studies (CADS) (Baker et al., 2008). CADS combines corpus linguistics with CDA while reflecting on critiques of the latter (Orpin, 2005). It mainly strengthens data representativeness and interpretative transparency of the analysis. CADS allows the investigation of the aggregate effects of language, highlighting typical discursive patterns. Czechia and Portugal represent intriguing study cases; the countries similar in area size, population, or GDP (Eurostat, 2024) differ significantly regarding their tech sectors and the length of the EU membership. Conceptually, this study approaches AI regulation visions with "sociotechnical imaginaries" (Jasanoff and Kim, 2015, p. 4), which have been plagued by conceptual ambiguity recently (Rudek, 2022). This research overcomes it by adopting a three-level imaginary concept (Sau, 2021). It structures the analysis of media representations by asking for imaginary's (1) social commentary, (2) vision of the future, and (3) means to achieve it. The research also highlights the roles of different related actors. Comparable corpora are collected from digitally available media in each country, covering the period of discussions about the EU's "AI Act" regulation (3/2018-12/2023). Results are explored and compared using Sketch Engine (Kilgarriff et al., 2014), analysing keywords, collocations and concordances. Such research provides innovative, empirically rooted comparative insights into the current media debate on the future of AI. Also, it provides a clearer perspective of sociotechnical imaginaries by grounding these to objective linguistic cues.