Digital Imaginaries and Public Opinion in Media Coverage of Global AI Governance

• In Event: Emerging Scholars Lightning Round: Graduate Student Research: Contemporary Issues in (Non)Democratic Governance

Abstract

This paper examines contemporary media portrayals of three major AI governance initiatives (namely, the EU AI Act, the US AI voluntary measures, and the Chinese Global AI Governance Initiative), to better understand how such narratives are associated with the general public’s beliefs around the future. By applying a digital imaginaries lens, we explore whether society is undergoing a technodeterministic shift in its perspectives on AI, or whether broader views of the future align with hopes and concerns about this technology in a more social constructivist manner. This sociological approach distinguishes our project from extant literature that approaches comparative AI governance from a policy or legal perspective.The study is divided into three components. First, we qualitatively analyse a series of news articles from reputable sources across the political spectrum and find common trends in coverage, expert sources, geopolitics, and emotionally-coded language. Second, we identify journalists’ use of digital imaginaries - portrayals of potential futures - and the meanings embedded within them. This is informed by the relationship between public opinion and journalist coverage. Third, we assess whether media framings are consistent with public opinion by comparing secondary public opinion data to the qualitative findings from the previous sections. The study results contribute to existing literature on media effects, whilst also nuancing ongoing public discourse on AI harms, existential risk, and AI governance.

Authors

• Hayden Goldberg, Oxford Internet Institute, Oxford University

• Will Shao, University of Oxford