Abstract News reporting on generative artificial intelligence (GenAI) and large language models (LLMs) plays a central role in shaping the public epistemology of these technologies, as journalism co-produces public agendas and narratives. Central to these debates are framing practices that portray LLM-based technologies as mentalistic, often through metaphors that attribute human-like cognitive capabilities or even experiences to AI systems (e.g. AI “thinks” or “feels”). While anthropomorphic framing has a long history in technology discourse, such practices carry particular social, cultural, and political weight when they contribute to inflated perceptions of AI capabilities. The present study analyses LLM-centric news framing with a focus on mentalistic representations, examining their thematic contexts, valence, and metaphorical patterns from a comparative, cross-cultural perspective. It combines computational text analysis with manual content analysis to critically explore 18,032 English-language news articles published between 2022 and 2024 across 15 outlets in Europe, Asia, and North America. The findings show that mentalistic framings are present but not uniformly dominant, varying across editorial cultures and journalistic styles. Situated against news media’s growing organisational and economic entanglement with GenAI, the study contributes to debates on AI imaginaries, public epistemology, and journalism’s role in fostering critical AI literacy.
Nguyen et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: