America’s AI Boom Widens Europe’s Investment Gap
U.S. business investment is projected to rise 40% from 2021 to 2027, against 12% in the euro area, as AI infrastructure concentrates capital and productivity gains.
The investment divide between the United States and the euro area is becoming a technology divide. New Oxford Economics forecasts reported by the Financial Times project that real US business investment in equipment and facilities will be 40 per cent higher in 2027 than in 2021. The equivalent increase for the euro area is only 12 per cent, while Germany is expected to record almost no growth over the period.
Artificial-intelligence infrastructure is a major reason for the divergence. Google, Meta, Microsoft and Amazon are together expected to spend more than $725 billion in 2026, much of it on data centres, chips, networks and the power systems needed to run them. That wave of capital expenditure is large enough to affect national investment data, industrial supply chains and productivity expectations.
The comparison is not simply between American software companies and European manufacturers. AI investment pulls demand through a wide physical economy: construction, electricity generation, cooling, semiconductors, fibre networks, specialised finance and engineering. When that spending concentrates in one region, the benefits can compound. Suppliers expand near major customers, skilled workers cluster around new facilities and lenders gain a deeper pipeline of financeable projects.
Europe does have important assets. It supplies advanced chipmaking equipment, industrial automation, power technology and high-value engineering. Several European regions also offer relatively clean electricity and cooler climates that can reduce data-centre operating costs. But those strengths have not yet translated into a comparable aggregate investment boom.
Oxford Economics has previously identified the constraints. AI activity remains smaller in Europe and relies heavily on imported technology. Planning and building approvals can take longer, while access to risk capital and to abundant, reliable electricity is uneven. The consultancy expects those limits to divert some projects toward the Nordic countries and southern Europe, but not to create a US-scale expansion in the near term.
A productivity problem, not only a spending contest
Investment matters because it equips workers and businesses with more productive tools. Financial Times reporting cited a sharp productivity divergence: output per hour worked rose by about $14 in the United States between 2018 and 2025, compared with roughly $2 in Europe. The figures are not a clean measure of AI's contribution, and they cover years before the current data-centre surge. They nevertheless show why policymakers care about the capital-spending gap.
For European companies, a slower investment cycle can raise the cost of adopting new technology. Fewer local data centres and compute providers may mean more dependence on foreign infrastructure. Smaller markets for growth capital can make it harder for software and industrial-AI businesses to scale before they are bought by larger overseas groups. The result can be a feedback loop in which weak investment reduces productivity, and weak productivity makes new investment harder to justify.
The gap also has consequences for financial markets. US technology groups are financing a growing share of capital spending from strong cash flows, but debt issuance, project finance and private credit are increasingly involved. Banks, infrastructure funds and energy investors can participate in that build-out. In Europe, the investable pipeline may be narrower even when institutional capital is available, because permitting, grid access and fragmented markets delay projects.
The $725 billion spending estimate should not be confused with proven economic value. Data centres depreciate, chips become obsolete and the revenue required to earn an acceptable return remains uncertain. If demand for AI services disappoints, the United States could be left with excess capacity and impaired projects. The same concentration that accelerates growth can amplify a downturn.
Europe's slower path therefore carries one limited advantage: it avoids committing as much capital before business models are tested. But caution is not a strategy if it also prevents productive projects. The relevant question is whether Europe is deliberately sequencing investment or simply failing to remove bottlenecks.
What Europe can still change
Closing the gap does not require Europe to copy every feature of the US model. It does require a faster route from capital to operating assets. That means more predictable planning, power-grid investment, deeper cross-border capital markets and procurement that gives young technology providers a credible first customer. It also means using Europe's industrial base as an advantage: AI applied to manufacturing, energy, pharmaceuticals and logistics may create more durable returns than a race to build consumer chatbots.
Regulation is part of the debate but not the whole explanation. The EU's AI framework adds compliance obligations, yet investment weakness predates the current rules. Energy costs, fragmented national markets, limited late-stage funding and slow project approvals are at least as important. Treating regulation as the sole cause would obscure the operational constraints that governments can address directly.
The forecast runs through 2027, so it is a scenario rather than a completed outcome. Corporate budgets can be cut, macroeconomic conditions can change and European investment could accelerate. The direction, however, is already visible: capital-intensive AI is widening a pre-existing transatlantic investment gap.
Why it matters
Business investment determines where infrastructure, supplier networks and future productivity are built. A 40 per cent increase in US capital spending against 12 per cent in the euro area would leave Europe more dependent on technology financed and operated elsewhere, even if European companies remain important suppliers.
For investors, the divergence creates opportunity and risk on both sides. US infrastructure offers growth but carries the possibility of overbuilding. European assets may look cheaper, but without faster deployment and stronger demand, valuation discounts can persist. For policymakers, the new numbers turn competitiveness from an abstract concern into a measurable capital-allocation problem.