| Summary: | <p>Artificial intelligence (AI) is increasingly being adopted in the news industry, but there are still few studies on the specific implications of AI; particularly on how it may reconfigure news organisations’ gatekeeping processes, reshape the news people get to see, and thus ultimately transform the public arena. This thesis addresses this gap. How AI potentially reshapes what news people get to see and the mechanisms of how they get to see it, the so-called gatekeeping process, will be a central component in understanding this topic. This has implications for the future of news organisations and, given the key role of journalism in shaping and representing modern societies, in reconfiguring the public arena and ultimately for democracy itself. To date, despite the importance of the topic, these questions remain largely unexplored.</p>
<p>With the aim of critically investigating these gaps, this thesis deploys a cross-case study of commercial and public service news organisations in the UK, the US, and Germany. I combine 174 interviews with news workers and experts as well as secondary data analysis, to answer the research question: <em>What are the structural implications of artificial intelligence for news organisations’ gatekeeping processes, and thus the public arena?</em> Under the umbrella of that research question, I examine three specific sub-questions: 1) How does the use of AI reshape the gatekeeping process of the news – in terms of production and distribution, 2) How does AI reshape the socio-technical organisation of news production and distribution in three national media systems, and 3) How does AI reshape the dependency of publishers on platforms companies?</p>
<p>After introducing the topic, I review the literature and theory and provide an overview of my methodology. The first paper then shows that AI is used in all parts news organisations to drive efficiency, optimise processes, and bring about greater effectiveness, reshaping the production and distribution of news by providing publishers with new means in the service of achieving certain existing ends, rationalising the work of news organisations, pushing it more strongly towards logics of efficiency, predictability, and calculability. The second paper presents a conceptual model of how and why the increasing use of AI in journalism reshapes the dependency of publishers on platform companies, leading to a loss of autonomy. I argue that platform companies are extending their control from distribution channels to the means of production through AI, limiting news organisations’ autonomy and their role as gatekeepers to the public arena on a structural level. The third paper provides further empirical support. I find that AI is indeed exacerbating the existing dependency of publishers on platform companies for distribution and creating new dependencies in production, with news organisations relying on platforms for AI due to high development costs, lack of resources, and differing visions of their mission. I conclude by discussing the individual and integrated contributions of each paper, the limitations of the thesis and provide ideas for future research.</p>
<p>At a moment when the news industry is in flux – reeling simultaneously from the ongoing erosion of its business models, increasing attacks from political actors, and various reckonings about its significance and changes in trust in news – it is vital to understand what AI will and can, and perhaps more importantly, what it will not and cannot do for and to the news. Similarly, in times when worries about the health of the public arena abound, the question of if, how and to what effect AI will reconfigure the news, the structural conditions of the public arena and thus deliberative democracy merits our attention. The adoption of AI in the news is bound to happen; in fact, it is already taking place. Understanding and theorising this development as well as its implications and informing decisions by contributing to our understanding of AI’s role in the news is the aim of this thesis.</p>
|
|---|