After the French AI unicorn Pigment was founded in 2019, it wasn’t long before investors began encouraging a move to the US.
“For many global SaaS companies, building a strong US footprint is essential to scale,” Eléonore Crespo, Pigment’s cofounder and CEO, says. “But I knew it wasn’t the right choice for me or for Pigment. Paris is where we started and it’s a city full of exceptional talent, especially in product, engineering and data; all of which are at the core of our company.”
It’s a strategy that’s paid off. Pigment now has a 500-strong team, the majority of whom are based in Europe. Last year, the company raised $145m in Series D funding.
“I think investors now see our approach is working,” Crespo says. “We have a strong, growing presence in North America [with offices in New York and Toronto] and that was always part of the plan. We just didn’t believe you had to give up your roots to do it. I’ve always believed we could build a truly global company from Europe.”
Pigment is one of Europe’s AI success stories but it’s by no means the only one. In 2024, European AI startups raised $8bn in funding, with standout rounds from France’s LLM startup Mistral (€600m Series B), Germany’s defence tech company Helsing (€600m Series D) and the UK’s autonomous vehicle startup Wayve ($1bn Series C).
Funding has continued to flow into the continent in 2025. At the recent AI Action Summit in Paris, 20 key investors (led by General Catalyst) announced plans to commit €150bn to European AI over the next five years.
Punching above its weight
One area where Europe punches above its weight in AI is technical expertise, says Andreas Klinger, founder of Prototype Capital, a first-round investor in European deeptech and hardtech.
The good news is the current cycle of startups favours deeptech and hard engineering.
Universities such as France’s École Polytechnique (alma mater to founders of Hugging Face, Dust, Mistral and others), along with the UK’s Oxford and Cambridge, and the Technical University of Munich, are helping maintain Europe’s high standards of AI research. According to analysis from Sequoia, Europe has a 30% higher concentration of AI experts than the US.
“American excellence is usually distribution rooted in a really fast growing, large market. That means really quick capital, really wide reach, scale, scale, scale. European excellence usually has its roots in engineering and research-heavy excellence,” says Klinger. He has invested in Dust, an enterprise AI assistant solution; Light Panda, which is developing a browser for AI agents; and Sensmore, a heavy machinery AI company.
“The good news is the current cycle of startups favours deeptech and hard engineering, and it’s a field where customers care less about where you’re based as long as you’re the number one supplier,” he adds.
What’s holding Europe back?
What’s holding Europe back from turning that AI expertise into global dominance? Part of the issue, Klinger says, is a lack of infrastructure.
“We have the right founders, with the right ambition but we need to fix the environment around them. Europe needs to be full-stack sovereign when it comes to AI and tech innovation — from data centres and semiconductors, to energy supply and the models themselves,” he says.
What is under-discussed, in some sense, is the full stack of value creation.
Estimates show that Europe will need to triple its energy capacity by 2030 to keep up with AI demand — in Ireland, that demand is already putting significant pressure on the country’s power grids, jeopardising its tech ambitions.
Europe also has a low market share for AI semiconductor design and manufacturing, cloud infrastructure and supercomputers, although there is progress being made. Mistral has committed several billion euros to build an AI data centre in France, and Nvidia has announced plans for at least 20 AI data centres to be built across Europe over the next few years.
“Europe’s AI industry is evolving rapidly but access to compute resources remains uneven,” says Roman Chernin, cofounder and chief business officer of AI-centric cloud platform Nebius. “The need for compute will only continue to grow. By making AI infrastructure more accessible, it ensures innovation is available to a broader ecosystem of startups and researchers.”
Nebius, headquartered in Amsterdam, was officially launched in 2024. The company has already raised €667m from investors including Nvidia and Accel, to expand its full-stack AI infrastructure, including large-scale GPU clusters and a cloud software platform. It operates data centres in Europe and the US, and recently announced a deployment of Nvidia Blackwell Ultra GPUs in the UK (the first in Europe).
“In terms of total capacity, we’ll be operating more than 100 megawatt by the end of the year,” Chernin adds. But he is quick to emphasise that Nebius is “not in the business of just building clusters or standing up data centres.”
“We’re here to deliver a robust cloud platform, with all the flexibility and performance you’d expect from a hyperscaler, but purpose-built for AI,” he says.
With new data centre projects in Europe being announced in rapid succession, Chernin warns that while physical facilities are great enablers, they do not solve the end task.
“What is under-discussed, in some sense, is the full stack of value creation, and it’s far beyond just physical infrastructure. At the end of the day, no matter how many gigawatts we build, someone needs to consume that compute and turn it into real value,” he says. “We’ve got some great teams, founders and products here, but the market is not as deep here as in the US. We need 100 times more.”
The global outlook
Beyond the data centres, supercomputers and energy capacity, Klinger says ambitious entrepreneurs also need a collaborative ecosystem that brings member countries together. He’s part of the EU Inc campaign, which argues the fragmentation of the continent is holding it back.
European entrepreneurs have every reason to be ambitious.
“We have no unified way to do IPOs in Europe, no good standards for early startup incorporation and 27-30+ times the amount of regulation when founders scale from country to country, compared to unified markets such as the US, China and India,” he says. “Europe has gaslit itself to believe it’s small. In reality, it’s one of the two largest players in the world.”
Crespo agrees and adds that rivalry with the US isn’t helpful for European AI founders.
“In AI especially, where the pace of innovation is fast and the competitive landscape is global, we have to think beyond our own borders — not just nationally, but continentally and internationally. A more harmonised internal market, especially for emerging tech, would give European startups the freedom and ability to grow faster and compete on a global stage.
“European entrepreneurs have every reason to be ambitious,” she adds. “We have the talent, the ideas and a growing ecosystem of support. What’s missing is the environment that allows those strengths to translate into scale. That’s what we need to fix.”
Read the orginal article: https://sifted.eu/articles/state-of-ai-europe-brnd/
