Europe must accept “ugly trade-offs” and build leverage in Artificial Intelligence rather than rely on moonshot projects that may not pay off, according to a group of European Artificial Intelligence researchers, think tank figures and investors. Europe 2031 is an 18,000-word warning about the cost of inaction, published on June 11. Framed around a fictional European Commission official trying to push Artificial Intelligence higher up the agenda, the scenario depicts a continent that is unprepared for the political will and capital mobilisation required.
The Commission presented its tech sovereignty package, which includes a proposal for a Cloud and Artificial Intelligence Development Act aiming to triple the EU’s data centre capacity within the next five to seven years. Data centres provide much of the computing power needed to train and run Artificial Intelligence systems. The Commission says expanding data centre capacity will require €200 billion, mostly from the private sector, by 2036. Meanwhile, the large-scale US cloud providers, known as hyperscalers, “are spending the same money in one quarter in 2026” on Artificial Intelligence data centres, Juijn said.
The proposed response focuses on leverage across energy, semiconductors, data centres, robotics and downstream applications. Europe can still support ambitious research into world models and other novel paradigms, including work associated with Yann LeCun, but Juijn said the bloc should operate under the assumption that it has lost the race to build today’s frontier large language models. Robotics and industrial Artificial Intelligence remain areas where Europe could compete, while partnerships with US hyperscalers may be needed as an intermediary route to capacity, despite ownership, sustainability and sovereignty concerns.
The sovereignty challenge remains unresolved because non-European providers may not be able to shield European data from third-country legal reach or service disruption. The Commission proposal would create four tiers of cloud and Artificial Intelligence sovereignty based on the perceived level of risk. The scenario warns that Europe could become trapped between US strength in cognitive Artificial Intelligence and Chinese dominance in physical Artificial Intelligence, with ASML among the strategic assets exposed to external pressure.
Recommended responses include a “middle-power coalition” linking countries such as the Netherlands, Germany and France with the UK, Canada, Japan and South Korea, as well as labour reform based on the Danish “flexicurity” model. The warning links sovereignty to enforcement of European values, noting that rules such as the Artificial Intelligence Act depend on foreign providers continuing to value access to the European market. If compute becomes scarce, US Artificial Intelligence providers may have less incentive to serve European users.
