Abstract

This article proposes and develops the concept of automated journalistic assemblages (AJAs). The concept unites the disparate types of algorithmic and Artificial Intelligence (AI) technologies created by high-tech and intermediary companies, newsrooms, and professionals in the journalistic field. It also draws together the current practical, research and regulatory debates and literature in the field to advance theories of media, journalism and technologies. The analytical framework informing the concept highlights at least three problematic dimensions in the current landscape of automated journalism, namely 1) the intensification of datafied-oriented news production and datafied-publics distribution; 2) the emergence of synthetic ‘churnalism’ with the popularisation of Large Language Models (LLM) and 3) the amplification of fabricated and inaccurate content. This article concludes that AJAs are useful for assessing the state of adoption of AI in newsrooms and aiding in its facilitation.