Europe may have fallen short in the AI labs, but it is succeeding in the toolshed.

      Europe may have missed the chance to establish AI labs, yet it is still benefitting quietly. According to Reuters, the continent's established tech and industrial leaders have unexpectedly thrived during the AI boom, even as startups struggle to compete with giants like OpenAI and Google.

      This situation illustrates where the financial rewards actually go. While frontier models attract attention, profits also flow to companies that produce the necessary tools, operate the software, and provide the energy, and Europe excels in all three areas.

      A prominent example is ASML, the Dutch company that manufactures the lithography machines essential for creating the most advanced AI chips, and which is nearing a trillion-dollar valuation as demand increases.

      Enterprise software serves as another key driver. SAP, now one of Europe’s most valuable companies, provides the systems into which large organizations are integrating AI, and every new AI feature it introduces encourages clients to remain loyal.

      Moreover, there is the less glamorous area of power and metal. Siemens has updated its forecasts regarding demand for AI-driven data centers, while electrical companies like Schneider and Prysmian are capitalizing on the necessary infrastructure that data centers require.

      The process is simple: every dollar spent on training a model eventually leads to chip purchases, and all advanced chips are processed through an ASML machine, meaning that expenditures which begin in California also find their way to the Netherlands.

      Telecommunications and utility companies also gain since data centers require connectivity and substantial energy, and European firms providing these services are meeting that demand, regardless of whether a single European model competes at a high level.

      The fact that this situation is considered surprising highlights a broader narrative. For the past two years, Europe has been portrayed as lagging in the AI race, with the continent evaluated solely on whether it has developed its own version of ChatGPT.

      By that standard, Europe did indeed fall behind and has expressed concerns about its AI sovereignty, relying on American models and cloud services for high-level tasks that its own companies are not yet capable of handling.

      There are exceptions, such as Mistral, the French company that has finally begun to realize its sovereignty ambitions, but a single lab cannot bridge such an extensive gap.

      The advantage for established companies is that they don’t need to lead in model development to profit. During a gold rush, selling tools has proven to be a more stable business, and Europe happens to control a significant part of that hardware supply.

      Collaborative initiatives are attempting to supplement the rest. A pan-European alliance has developed an open LLM as an alternative to American and Chinese options, though it remains a modest competitor to the larger entities.

      While the advantages are real, they are also limited. They depend on selling products during a boom that incumbents do not control, and a slowdown in AI development would consequently decrease demand for their tools and power.

      This situation has also altered Europe’s corporate landscape. The continent’s most valuable businesses are increasingly its industrial and tool manufacturing firms related to AI, rather than financial institutions or luxury brands.

      The gains are also unevenly distributed, concentrated in chip tooling, industrial software, and electrification, while Europe’s consumer internet and social media platforms remain largely disconnected from the AI narrative.

      For policymakers, this outcome is problematic. Europe is profiting from AI while still relying on others for the core technology, a comfortable yet precarious situation.

      The risk lies in misinterpreting the profits as a strategy; selling into someone else’s boom can be profitable until that buyer develops its own supply chain, and Europe’s reliance on external models represents a weakness that a strong quarter cannot resolve.

      Nonetheless, acting as the provider for a global boom is not the worst position to be in. While Europe may not have created the AI future, it is selling a significant amount of what supports that future.