AI Pluralism Monitor: First Country Databases Roll Out from June to September 2026

The Media and Journalism Research Center is launching the first country databases of its AI Pluralism Monitor, a new project that maps how artificial intelligence systems make news sources visible, cite them, rank them — and sometimes leave them out.

The databases roll out progressively between June and September 2026. Rather than a single release, the first phase is published country by country as each baseline is built, tested across major AI systems, and coded. The opening phase covers ten countries, Romania, Hungary, Serbia, Germany, Spain, the United Kingdom, Brazil, India, South Africa and the Philippines, chosen to capture variation in language, media-system structure, platform dependence and media-capture risk.

As people increasingly turn to AI assistants and answer engines to ask about politics, current affairs and media, these systems are becoming intermediaries between citizens and the news. The AI Pluralism Monitor tests what they surface, how transparently they attribute it, and whose journalism they make visible across nine major AI systems and two language layers, English and the main local language.

Each country database is accompanied by source-visibility findings and an open methodology, including the prompt bank and coding framework, so results can be scrutinised, replicated and expanded. The database is designed to grow beyond the first phase to include further countries, regions and language environments.

Explore the project: AI Pluralism Monitor →