Navigating artificial intelligence narratives in Chinese media: a mixed-methods discourse analysis

Abstract With the deepening social penetration of artificial intelligence (AI) technology, mainstream media’s role in shaping public cognition through news framing has become increasingly prominent. However, existing research has long been constrained by Western-centric perspectives, leaving significant gaps in exploring the discourse construction mechanism related to AI in non-Western contexts. This study seeks to systematically deconstruct the mainstream media narrative strategies on AI technology in the Chinese context, by employing a hybrid methodology combining Analysis of Topic Model Networks, Moral Foundation Analysis, and linguistic analysis and using AI-related news reports from 2017 to 2024 in Chinese mainstream media as a corpus. The findings suggest that, first, the media have constructed four frames of “Politics and Economy,” “Society,” “Revolution,” and “Cooperation,” whose structural details reflect the deep coupling between technological discourse and national strategies. Second, moral discourse demonstrates Chinese characteristics prioritizing dimensions of Authority/Subversion and Loyalty/Betrayal, whose diachronic growth correlates with contextual factors such as pandemic governance and nationalist reinforcement. Third, through the combined linguistic strategies of predication and metaphor, media narrate AI as a manipulable national development tool while downplaying technological complexity and threats. This study not only reveals the mutual-construction mechanism between ideology and discursive strategies in technological socialization, but also provides Chinese empirical references for global AI governance research through a non-Western perspective.