Europe doesn't need another Silicon Valley to succeed
Investor Andrew Gershfeld argues that Europe should focus on its strengths in AI rather than trying to imitate Silicon Valley.
Europe's choice: specialize or follow the US model
Since AI became mainstream, comparisons between Europe’s innovation hubs and Silicon Valley have become ever-present. In 2026, the EU was still debating domestic cloud capacity, Big Tech dependence, and the energy demands of new data centres. Investor Andrew Gershfeld from Flint Capital emphasizes that Europe does not need to win every layer of the AI stack; it should specialize in areas where it already has advantages.
The pull of the US market and challenges for European startups
Capital, customers, acquirers, and commercial talent still concentrate in the US. While seed and Series A capital in Europe is more achievable than a decade ago, growth rounds often require a stronger US connection. Europe lacks enterprise software companies that repeatedly sold into large US customers at scale, so the pool of executives with that go-to-market experience is limited. If a company’s revenue is still European, investors will continue to see it as European, regardless of its competitor list.
Where Europe genuinely wins
Defence and security technology are good examples. Although the US market is larger, procurement rules, nationality questions, and compliance requirements can make entry difficult. Industrial software is another area where Europe has an advantage. A startup building AI for logistics, manufacturing, energy, healthcare, or industrial maintenance benefits more from expert customers than from proximity to large venture funds.
Practical checks and conclusions
When a European founder asks about US expansion, the investor first checks who they need to sell to, who can sell it, and whether the company has enough US proof to support its fundraising story. Europe’s task is to remove barriers that hinder founders, such as taxing paper wealth or complex hiring regulations. Europe does not need to win every layer of the AI stack – it can win where customer intimacy, regulatory understanding, engineering depth, and trust matter more than raw compute.
