Artificial Intelligence in the Public Eye: The Social Construction of AI as a Problem in the National English-Language Canadian Newspapers

This chapter examines how national English-language Canadian newspapers have constructed Artificial Intelligence (AI) as a public problem in the wake of the release of ChatGPT. Drawing on the representational naturalization model (Negura & Plante, 2021) and social problem theory, the study analyzes 208 articles published in The Globe and Mail and National Post between November 2022 and November 2024. Using a mixed-methods approach—topic modeling, sentiment and emotion analysis with pre-trained language models, and qualitative thematic coding—the study identifies five major themes: job substitution, applications of large language models in education and healthcare, multimedia generation and deepfakes, legal regulation, and economic impacts. Findings show that media discourse is predominantly problematizing, marked by negative sentiment and dominated by emotions such as fear, anger, and sadness, particularly in relation to job loss, privacy violations, and ethical risks. Positive narratives persist, especially regarding economic growth, reflecting ambivalent and polyphonic representations. The analysis demonstrates how sensational incidents amplify collective anxiety and accelerate the framing of AI as a risk-laden social issue. The study contributes to understanding AI’s public construction by highlighting the cyclical and affective dynamics of media discourse, where emotions function as catalysts of problematization, mobilization, and demands for regulation.