AI for Democracy? How Public Service Media Around the World Use Artificial Intelligence to Produce Journalism

We examine how public service media (PSM) around the world use AI to assist journalism production. Using surveys, organizational documents, and interviews with senior managers from PSMs in Africa, Asia, Europe, Oceania, and the Caribbean, we map which AI tools PSMs use, how they use them, and how they navigate the relationship between efficiency, PSM values, and their broader democratic purpose. We analyze participating PSMs’ use of AI in transcription, image generation/editing, and content customization for online and social media. Other uses, including translation, synthetic voices, and Retrieval-Augmented Generation, AI “presenters,” and facial recognition, are also discussed. We argue that participants’ uses of AI, and the tools they deploy, are not particularly distinctive: a key driver is also time and cost savings. However, we argue that the use of AI by these PSMs does involve uniquely complex challenges relating to marked tensions between efficiency objectives and PSM values, between PSM values themselves, and between PSMs’ efforts to fulfil their statutory obligations to audiences and their broader democratic mission. This study provides the first broadly-based international study on PSMs’ use of AI, highlighting the difficulties they face in safeguarding democratic journalistic values despite rising costs and budget cuts.