Europe's AI Ambitions Hinge on Corporate Buyers, Not Just Venture Capital

Money is flowing into European artificial intelligence ventures at unprecedented rates, yet capital alone cannot secure the continent's technological sovereignty. The real bottleneck, conversations at the HumanX conference in Amsterdam revealed, is a shortage of local customers willing to deploy what these startups create.

In the first half of 2026, European AI-focused startups pulled in $23 billion, a 130% jump from the prior year and representing 55% of all venture funding in the region, according to joint research from Crunchbase and HumanX. This surge has intensified debate over how nations should allocate resources to capture AI's economic benefits and maintain sovereignty over their data and systems.

Yet building every component of the AI stack internally is neither necessary nor realistic, according to two technology leaders who spoke at the conference. Fabrizio Del Maffeo, founder and chief executive of Axelera AI, and Mehdi Ghissassi, chief product and technology officer at AI71, outlined how strategic choices about specialized hardware, applications and purchasing power can shape whether AI investments translate into genuine competitive advantage.

The Five-Layer Model

Nvidia chief executive Jensen Huang has outlined a "five-layer cake" architecture for AI systems: energy, chips, infrastructure, models and applications.

Del Maffeo sees opportunity in the semiconductor tier, particularly as artificial intelligence computation shifts away from centralized facilities toward edge devices. "Artificial intelligence will expand from cloud computing, from centralized data centers, to devices closer to us in the physical world," he explained. "To enable this, you need specific chips which can run efficiently, at a lower cost, to connect these networks that today are running in the cloud. We are here to solve this problem."

Axelera AI, established five years ago, has developed two generations of inference chips and serves approximately 600 customers. The company intends to move into products serving decentralized cloud infrastructure in coming months.

The Application Layer Matters Most

Abu Dhabi-based AI71 works alongside major corporations and government bodies to guide them through AI adoption and transformation.

The United Arab Emirates possesses the world's highest compute capacity per resident and has mandated that all government agencies establish AI-driven processes for citizens within three months, Ghissassi noted. Abundant, inexpensive energy supports this infrastructure. Over the next twenty-four months, the target is to handle half of all citizen interactions with government through AI agents.

Ghissassi argued that competing in the model layer makes little sense. Only a handful of companies—perhaps two, three or four—can sustain the enormous capital expenditure required, and the field commoditizes rapidly. The application layer, by contrast, is where sovereignty truly matters. "You want to make sure that it stays with you, be it that you're a government or an enterprise," he said, drawing on his prior experience leading product work at Google DeepMind. "If you're giving away your trade secrets and know-how, nobody stops whoever is being a provider to you today, from replacing you."

In the UAE, success stems from "the mandate, the pace at which things happen, the availability of compute, both in terms of sovereign clouds, or on-prem, or global clouds, and then the amount of capital that is being invested to help transform these companies to benefit from these technologies," Ghissassi said.

Europe's Structural Disadvantages

Europe faces structural headwinds. As a net energy importer, it lacks the cheap power that fuels AI infrastructure. While the continent boasts established semiconductor expertise, it is starved for the computational capacity needed for neural network training and frontier research labs.

"We should not be obsessed with controlling the entire stack," Del Maffeo said. Instead, Europe must focus on creating economic value rather than simply purchasing services. "Creating value means creating a wealthy economy."

The region possesses genuine strengths: deep research talent and a population of 440 million potential beneficiaries of AI technology. However, European enterprises have not developed the same appetite for buying from startups that characterizes the United States, limiting the growth momentum that such relationships generate.

"What worries me is that we are a little bit lagging behind, and therefore we are missing this value creation, and this will weaken the economies of Europe," Del Maffeo warned.