Giving AI a Body: Representations of Social Robots in Latin American Media

Abstract

Social robots (physically embodied systems designed to interact with humans through social behaviours) have gained increasing visibility in news media. However, their everyday presence in Latin America remains limited, reflecting broader asymmetries in technological diffusion across non-WEIRD contexts. This study examines how social robots are represented in Latin American news media and how artificial intelligence (AI) is communicated through embodied technologies. A quantitative content analysis was conducted on a corpus of 404 news articles published in Spanish and Portuguese language media, retrieved through Google’s search engine. The study employs an LLM-assisted coding procedure combining automated classification with human validation. The analysis focuses on three dimensions: journalistic treatment (narrative frames and evaluative tone), representation of AI (attributed agency and explicit references to AI), and social functions (social roles, sectors of application, and sources). The findings show that coverage is predominantly framed around care and assistance, with positive or ambivalent evaluations, while conflict-oriented frames remain marginal. Social robots are mainly portrayed as assistants or caregivers, described as semiautonomous systems, and situated in healthcare and research contexts. Coverage relies largely on corporate and scientific sources, with a limited presence of civic or governmental voices. These representations position social robots as a key interface through which AI becomes publicly interpretable. The study contributes to research on AI and journalism in non-WEIRD contexts by showing how media representations shape understanding of emerging technologies in the absence of widespread direct experience.