In her 2026 State of the Union Address on 16 September, European Commission President Ursula von der Leyen revealed where she thinks Europe's real advantage in artificial intelligence lies. Europe does not have to build the frontier models to gain the most from AI. The value comes from putting AI to work in the real economy: on the factory floor, in energy grids, in agriculture and in healthcare.
Her argument is based on one high-quality asset that current European companies holad: data. She described it as Europe's great untapped advantage, the raw material for tailor-made industrial AI models.
“These data are a precious European treasure. They are what can power tailor-made industrial AI models.”
— Ursula von der Leyen, State of the Union address, 16 September 2026
She named advanced manufacturing as one of five priority sectors, including health, transport, agri-food, defence and space.
The advantage only works if the data can move
That advantage comes with a condition. Industrial data becomes valuable when it can be combined across companies and along a value chain. A single machine log or quality record has limited use on its own. Its value grows when it can be shared and analysed together with data from suppliers and partners.
This is where many manufacturers stop. Sharing data raises real concerns about control, confidentiality and commercial risk. If a company cannot share data on terms it trusts, the industrial AI that von der Leyen described stays out of reach, however good the underlying models become.
Trust is the missing piece
Von der Leyen was clear that Europe should do this in its own way, with human oversight at the centre. She made the point through healthcare: AI should support a doctor's diagnosis, not take the doctor's place. The same holds in industry. Companies want the benefits of shared data without giving up control over how it is used.
Trusted data cooperation is what makes this balance real. Approaches built on principles such as those promoted by Gaia-X let organisations exchange and use data across companies while keeping control over access and usage. This is the groundwork any industrial AI ambition depends on.
VET4DATACOOPERATION, an Erasmus+ KA220-VET project, is building vocational training to help manufacturing companies develop this capability: the knowledge and practical skills to share data with confidence. Europe's data treasure is a real high-quality asset, but turning it into value depends on knowing how to cooperate around it.
Visit the official LinkedIn page for VET4DATACOOPERATION project updates: https://www.linkedin.com/showcase/vet4datacooperation/posts/
