While estimates of AI spending across regions are hard to find, every available measure shows the scale of investment in Europe is an order of magnitude smaller than in the US and China.
To take one example, AI-related capex in the US is likely to exceed 2% of GDP this year, while the equivalent figure for the EU remains a small fraction of that. On AI infrastructure, one estimate has the US accounting for roughly three quarters of the global aggregate, with China a distant second and Europe well behind both.
In a previous paper on AI in Europe, we outlined reasons for historically lower levels of IT investment, including sector composition and management practices. But two additional bottlenecks are slowing the AI buildout: higher energy prices compared with the US, which makes investment in power-hungry data centres less attractive, and an aging electricity grid that lacks the capacity to connect them.
That said, Europe is substantially exposed to AI spending and will share in gains and losses from the capex boom, even when the physical infrastructure is built elsewhere.
Europe is also a key player is the semiconductor supply chain, where EU firms are deeply embedded in the production process that makes AI models possible. ASML is the best-known example and its equity returns continue to correlate very strongly with US tech. Although the company’s capex spend is a fraction of the US tech giants, advanced chip production remains dependent on lithography equipment that only a handful of firms can provide and where ASML has maintained a dominant position.
For some time, the emphasis in European policy was on regulation and safety, but that has shifted towards encouraging adoption. Discussions in the Draghi Report and subsequent European Commission initiatives acknowledge Europe's investment and innovation gap, yet progress on boosting domestic investment through reforms has been slow and the sums raised are still small.
Instead, the focus is on strengthening the conditions that allow European firms to adopt AI more quickly through investments in computing access, digital infrastructure, skills, energy availability, research collaboration and capital-market development. The underlying strategy is therefore not to compete with the US and China on capex levels, but to accelerate adoption.
