The good, the bad, and the obscure: Exploring uses of generative AI in the global 2024 elections

Digital technologies are ubiquitous in political communication, but their democratic consequences remain contentious, with the recent emergence of generative AI (GenAI) further complicating already complex sociotechnical issues. This article explores the empirical terrain of GenAI uses that came to public attention in the 2024 global elections by collating and analysing data from two prominent tracking projects. First, we map where GenAI was identified, which actors were involved, and through which platforms and modalities it circulated. Second, we build a provisional typology of emerging GenAI use in election campaigns, identifying practices of informing, campaigning, dissenting, disinforming, defaming, impostering, and spoofing as a basis for discussing the in-/authenticity of communicators and the in-/correctness of content. The article nuances proposed assessments of the political implications of generative AI by offering one of the first systematic, empirical analyses of its potential democratic consequences, identifying context-specific good, bad, and obscure uses of these new technologies.