Abstract
This paper explores the integration of artificial intelligence (AI) in investigative journalism for scrutinizing AI algorithms behind large digital platforms. It draws lessons from two journalistic case studies that leveraged AI to analyze AI models from large digital platforms: the scrutiny of training data for large language models and the analysis of TikTok's recommendation algorithm in the context of eating disorders. Through practical examinations, the paper outlines three ways in which AI techniques can assist in navigating the technological complexity of large digital platforms: speed and scale, personalization, and reproducibility. Furthermore, through the case studies presented in the paper, the study aims to show how small interdisciplinary teams of journalists and data scientists can effectively hold digital platforms accountable and present complex algorithmic processes in an accessible manner to the public.
Original language | English |
---|---|
Number of pages | 19 |
Journal | Journalism Practice |
DOIs | |
Publication status | E-pub ahead of print - 19 Mar 2025 |
Bibliographical note
Publisher Copyright:© 2025 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
Keywords
- Algorithmic accountability
- Artificial intelligence
- Computational journalism
- Data journalism
- Digital platforms
- Investigative journalism
- Platform observability
- generative AI