Europe’s public broadcasters want artificial intelligence. They also want to keep their data on their own turf. Nvidia is offering to help them do both.
The chipmaker and the European Broadcasting Union (EBU) have partnered to build sovereign AI and hybrid cloud frameworks for European public service media. The deal was announced on June 11, 2025, at Nvidia GTC Paris during VivaTech. It covers a network that collectively reaches an audience of more than 1 billion people.
What the partnership actually involves
The EBU counts over 110 member organizations spread across more than 50 countries.
The stated goal is to deliver trusted technology that meets European standards on data protection and privacy.
The EBU will integrate two of its architectures: the Dynamic Media Facility (DMF) and the Media eXchange Layer (MXL). Both are designed to support efficient interoperability and AI workflows across member organizations.
Nvidia’s contribution includes the GB200 Grace Blackwell Superchip, its Holoscan for Media software, and DGX Cloud.
Nvidia will offer training through its Deep Learning Institute, so member organizations can build the skills to actually use these tools.
The partnership also lists interoperability, scalability, and energy efficiency as priorities across the EBU network.
Why broadcasters asked for this
This collaboration responds to earlier calls from the EBU Technical Assembly for cloud and AI infrastructure that is resilient, sustainable, and interoperable.
The two organizations are expected to hold multiple joint sessions and presentations through late 2025 and into 2026. Those events are set to focus on sovereignty in AI for broadcasting, including the ongoing integration of Holoscan for Media with EBU architectures.
What this means for Nvidia and European media
For Nvidia, the deal fits a clear pattern. The company has spent a growing share of its public messaging on “sovereign AI,” the idea that nations and regions want computing capacity under their own control.
The EBU deal gives that pitch a concrete use case, pointing to a network of over 110 media organizations adopting a shared framework built around it.
For broadcasters, the upside is a common technical foundation. Members operating in more than 50 countries often face very different budgets and capabilities. A shared framework, plus training through the Deep Learning Institute, could help smaller organizations access tools they would struggle to build alone.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.