Europe Doesn’t Need Another AI Strategy. It Needs Buyers.

Europe is investing in AI infrastructure, developing European models and discussing how to become more technologically independent. I welcome that. But a European technology company cannot survive on political ambition alone. At some point, someone needs to sign a contract.

Whenever I read another announcement about Europe’s AI ambitions, I find myself asking a simple question. Who is going to buy what we build?

Europe has become good at funding AI innovation

Europe has no shortage of talented engineers, researchers or entrepreneurs. Our universities do excellent research, companies are building sophisticated technology and public programmes are designed to support them.

The European Commission’s AI Continent Action Plan aims to mobilise €200 billion in AI investment. The infrastructure being developed could make a real difference to European companies.

But access to computing power is only part of the equation. You can give a company funding, access to infrastructure and introductions to the right people. None of that replaces a paying customer.

A customer forces a company to solve a real problem, make its product reliable and deal with integration, security, support and the thousand practical details that never appear in a pitch deck. Most importantly, a customer gives the company revenue it can use to build the next version.

A pilot is not a customer

A lot of energy goes into preparing technology demonstrations. The technology works. People in the room are interested. There is genuine enthusiasm about what might become possible.

Then comes the difficult part. Who owns the budget? Who can approve the purchase? What procurement process applies? Can the organisation actually deploy the software?

A successful demonstration can be the beginning of a business relationship. It can also become another demonstration six months later.

For a young company, the difference matters enormously. Engineers need to know whether they are building a product that will be used or preparing another presentation for a committee.

I would rather see a European technology company win one demanding operational customer than collect ten enthusiastic letters of intent.

Public procurement is an industrial policy decision

This matters particularly when the customer is a public institution. Governments are among the largest buyers of technology. Their purchasing decisions influence which companies grow, which products mature and which capabilities remain available in Europe.

Of course, public authorities must spend responsibly. They need security, reliability, transparency and proper competition. I would not argue for buying an inferior product simply because it comes from Europe.

But I do think we should ask whether our procurement systems give capable European companies a realistic opportunity to compete.

Can a smaller provider meet the requirements without employing an entire department dedicated to tenders? Can a public organisation purchase a limited operational deployment without committing to a massive programme? And when a European company demonstrates that its technology works, is there a credible path from that demonstration to a contract?

These are practical questions. They deserve as much attention as the next funding announcement.

The first serious customer changes everything

There is a particular kind of progress that happens when software enters daily operations. People stop discussing what the product could theoretically do. They start telling you what is missing.

The interface is too slow. A workflow needs fewer steps. A data source behaves differently than expected. A feature that looked impressive in a demonstration turns out to be less important than something nobody mentioned before.

This feedback is valuable because it comes with consequences. Someone depends on the system working tomorrow morning.

That pressure improves products in ways that workshops and strategy papers rarely can. It also creates something investors, partners and future customers understand. Evidence that the technology is useful.

Europe should be willing to become its own customer

The European Commission’s Apply AI Strategy recognises the importance of adoption and encourages a European purchasing approach, particularly in the public sector. That is a sensible direction.

But the real test will happen inside individual organisations.

Will purchasing departments create workable routes for innovative suppliers? Will decision makers accept the responsibility of choosing a newer European product when it meets the requirements? Will established companies give smaller technology providers a genuine opportunity to prove themselves?

And will European technology companies be prepared to deliver the reliability, support and accountability those customers have every right to expect?

The responsibility runs both ways. European companies cannot demand contracts as a reward for being European. They have to earn them. But they also need customers willing to evaluate what they have built rather than dismissing them because they lack decades of references.

The next European AI success story needs a purchase order

I want Europe to invest in AI infrastructure. I want us to develop competitive models and attract excellent engineers. We need all of that.

But I would like to see us become equally ambitious about buying and deploying the technology our companies create.

Not another innovation showcase where everyone agrees the future looks promising. An actual system, used by actual people, solving an actual problem. With a customer who pays for it, expects it to work and helps make it better.

That is how technology companies become strong. And it is how technological sovereignty becomes something more than a policy objective.

Europe has spent enough time asking who will build its AI future. I think we should start asking who is prepared to buy it.