VRT strengthens European collaboration on transparent and trustworthy AI

Thursday, 19 June, 2025 – The European Commission has launched ELLIOT, an ambitious research and innovation project aimed at accelerating the development of open and trustworthy AI models in Europe. VRT, together with KU Leuven, will participate as Belgian partners in this four-year initiative, which is funded through the Horizon Europe programme. In addition, VRT is joining forces with the Flemish branch of OpenChip, an organisation working on a European alternative for secure and sustainable AI chips.

AI for Europe

The project brings together 30 organisations from 12 European countries. The goal? To develop powerful AI systems capable of handling various types of data, such as text, images, video, sensor data, and satellite imagery. These 'multimodal' models would be able to generalise and reason better than the current generation of AI and are designed to operate in today’s complex, dynamic reality.

What makes ELLIOT unique is its emphasis on openness and collaboration. All models will be shared openly, making them widely applicable – from media and mobility to climate research and public services. Thanks to Europe’s most powerful supercomputers, such as the Barcelona Supercomputing Centre, these models can be trained at scale.

Media tailored to the public

Stijn Lehaen, General Director of Technology & Infrastructure at VRT: “AI applications are only as smart as the data they are trained with. The richer the data, the better the models will perform. With this project, we are laying the foundation for AI applications that the users of tomorrow can fully trust. As a public broadcaster, we consciously deploy technology in service of society, where innovation always goes hand in hand with our public values.”

For VRT, participating in ELLIOT marks an important step in exploring AI applications within the media sector. Think of smarter search functions, automatic subtitling, or personalised chatbots – all developed with respect for European values such as transparency, privacy, and inclusion. VRT is also investigating how AI can contribute to a reliable news offer, for example by detecting disinformation.