Journalism in the Age of Generative AI: A Theoretical Framework of Newsroom Automation, Editorial Authority and the Economics of Trust

Abstract

Generative artificial intelligence has entered newsrooms faster than any previous technology. Reporters use it to transcribe, translate and summarise; some outlets publish text, images and even presenters produced by machines; and platforms use related systems to decide which stories reach which readers. Research on these developments has grown rapidly, yet it remains fragmented. Studies of newsroom adoption, of audience responses to AI disclosure, and of platform distribution rarely speak to one another, and most draw on Western and East Asian cases. As a result, there is no framework that explains how automation at different points in the news process redistributes editorial authority and alters the economic value of trust, especially in media systems such as Nigeria's, where trust is scarce, newsrooms are under financial strain and journalists face political pressure. This conceptual, non-empirical paper addresses that gap. It integrates peer-reviewed scholarship published between 2023 and 2026 on generative AI in journalism, audience trust and AI disclosure, algorithmic news distribution and platform power, and Nigerian and African journalism. Every source was checked against its Crossref record and screened for retractions and editorial notices. From this synthesis, the paper develops a production–generation–targeting framework. Production refers to AI assistance in newsgathering and editing; generation refers to AI that creates publishable content; targeting refers to AI that personalises, ranks and monetises distribution. The analysis shows that each layer moves editorial authority to a different place: production shifts it towards tool vendors and workflow design; generation shifts it towards models and the people who prompt and check them; and targeting shifts it towards platforms and recommendation systems. It also shows that trust behaves as an economic asset whose value rises as synthetic content spreads, even as AI disclosure tends to lower perceived trustworthiness in the short term. Five propositions follow, concerning substitution under financial strain, the disclosure dilemma, the platform capture of audience relationships, the amplification of existing trust deficits, and the conditions under which editorial authority is retained. The paper concludes with recommendations for Nigerian newsrooms, regulators, journalism educators and platforms.