This article examines how topic diversity in Norwegian news media varies across proprietary and social media platforms. Drawing on a comprehensive dataset of 224,905 articles published in 2023 by 22 Norwegian news outlets, and corresponding posts on Facebook, Instagram, and TikTok, the study employs Latent Dirichlet Allocation (LDA) topic modeling to identify and compare thematic patterns. Results reveal systematic adaptations to platform-specific logics: while the outlets’ online news sites maintain broad topical diversity, they favour politics and human-interest stories on Facebook, personalised and entertainment-oriented content on Instagram, and sensationalist, event-driven topics on TikTok. Statistical diversity measures confirm that the outlets’ social media posts display lower topical diversity than their online articles. These findings demonstrate that editorial strategies align with distinct platform logics, contributing to a segmented news environment. The study argues that such differentiation has implications for the democratic role of journalism, as audiences encounter increasingly platform-tailored representations of news.
