The EU wants to team up with the US on AI

Europe wants its own AI sovereignty, but it also knows it can't afford to get left behind by Silicon Valley.

During a G7 session on artificial intelligence on Wednesday, European Commission President Ursula von der Leyen publicly urged the United States and the European Union to forge a much tighter alliance on frontier tech.

Her main selling point? Scale. Von der Leyen pointed out that together, the US and Europe control roughly 70% of the entire global AI market.

A transatlantic pact for the AI era

Posting on X following the session, von der Leyen called AI an exponentially growing shift and the defining technology of our era, pointing out that it brings massive upside alongside serious risks to democratic societies.

She argued that Europe and the US share complementary strengths and security interests, meaning the two powers have a responsibility to lead together.

Her roadmap calls for three concrete moves: pooling capital to invest jointly, accelerating real-world deployments in fields like manufacturing and healthcare, and putting strict guardrails on the most capable frontier models so they stay safe and trustworthy.

Sovereignty fears lurk beneath the surface

Despite the friendly tone at the G7, the call comes right as European leaders get increasingly anxious about over-relying on American tech giants.

Earlier, the European Commission explicitly warned about this dynamic after Anthropic curbed access to its advanced AI models for users outside the United States. That rollout restriction served as an immediate wake-up call in Brussels about technological sovereignty and foreign platform dependency.

Mistral AI CEO Arthur Mensch recently voiced a similar urgency, warning that Europe has only a narrow window to stand up its own independent compute and model infrastructure before it falls into irreversible dependence on US providers.

He pointed out that the global AI race is ultimately governed by access to chips, raw power, and compute, while European players are continually held back by heavier regulation and smaller capital pools.

Even Wall Street sees the tension. JPMorgan highlighted Mistral as Europe's strongest native contender in a market otherwise dictated by the US, predicting a sharp rise in corporate and governmental demand for "sovereign AI" as organizations fight to keep their sensitive data on home turf.

Why it matters

We think von der Leyen's pitch is a pragmatic attempt to buy a seat at the steering wheel of frontier AI.

By framing the relationship around a shared 70% market share, Europe is trying to ensure it doesn't get relegated to a mere consumer or regulator of American models. For enterprises operating on both sides of the Atlantic, closer alignment on deployment and safety benchmarks could eventually mean less regulatory whiplash between Brussels and Washington.

That said, the tension between transatlantic cooperation and local sovereignty isn't going away.

It looks like Europe wants the benefits of America's hyper-advanced frontier models while actively funding homegrown champions to avoid getting locked out, as Anthropic's regional limits demonstrated. Until Europe solves its own compute and capital bottlenecks, calling for an equal partnership will remain easier said than done.