German Broadcasters Test Decentralized AI Data Ecosystem

German media companies are initiating a project to test the foundations of a shared data infrastructure. Broadcasters ARD, ZDF, RTL Deutschland, and ProSiebenSat.1 are collaborating with the German Press Agency (dpa) to develop the technical and organizational foundations for the Media Data Space, a national media data infrastructure aimed at creating a decentralized, AI-integrated data ecosystem.

The initiative seeks to enable organizations to independently enhance digitalization programs amid global competition. Specific use cases from the media sector are being tested to strengthen the infrastructure's foundations. The project focuses on authenticity, deepfake detection using shared AI models, harmonizing data usage, content personalization, and fine-tuning AI language models.

Data control remains with the respective participating partner at all times under the 'compute to data' principle, with AI 'traveling' to the data. After the test period, a decision will be made on establishing a permanent national Media Data Space, starting in 2026.

Evaluation of the results will occur in Munich on October 16th. This marks a significant step in the digital transformation of the German media industry.