The AI Factory

Luxembourg has launched something called the AI Factory. Economy Minister Lex Delles announced it at Luxinnovation's annual review with a statement that doubles as a mission statement and a warning: "Innovation is no longer an option, but a necessity for every company."[1]

The AI Factory, developed by Luxinnovation, is designed as a central platform, a "one-stop shop" to guide companies through every stage of an AI project, from initial exploration to deployment. Its centrepiece is a service catalogue with 80 structured offerings.

Eighty services. That is a lot. It is also the first sign that this might be more bureaucracy than breakthrough.

What it actually does

Luxinnovation handled over 2,000 requests last year, provided direct support to more than 560 companies, and helped mobilise over 50 million euros in European funding for the Luxembourg economy. About 85% of the businesses they supported were SMEs. Those are solid numbers. The agency is not starting from zero.

Mario Grotz, Luxinnovation's director, predicts that AI will fundamentally transform business models across all sectors. "Soon, there will be hardly any company without AI," he said. He also emphasised preparing businesses gradually, making them aware of both opportunities and risks. "If used correctly, AI can drive efficiency, but it is also important to understand the risks before implementation."[2]

This is all reasonable. But it is also the kind of reasonable that becomes a trap.

The catalogue problem

An 80-service catalogue is not a product. It is a menu at a restaurant where nothing is cooked yet. The AI Factory's own description acknowledges this: "It is not about explaining all 80 services, but about listening to what companies require, and then proposing" the right ones. Translation: we have built a framework so broad that we need to explain why breadth is not the same as depth.

The companies that need AI the most, small manufacturers, local service businesses, retailers running on spreadsheets, do not need 80 options. They need one thing that works. They need someone to sit with them, understand their process, and show them how a specific model or tool saves them actual hours. A catalogue is what you build when you do not know who your customer is yet.

The real signal

What is genuinely interesting here is the positioning. Luxembourg is a small country with a big financial sector and limited industrial diversity. An AI adoption agency that focuses on SMEs rather than banks is betting that the next wave of AI value creation happens in boring, practical, operational improvements, not in trading algorithms or compliance automation.

That bet is probably right. The companies that will benefit most from AI in the next five years are not the ones building foundation models. They are the ones figuring out how to use them. A small accounting firm that automates invoice matching, a logistics company that optimises route planning, a bakery that predicts daily demand from weather data. These are not glamorous use cases. They are the ones that add up.

If the AI Factory can resist the temptation to become a consulting firm with better slides and instead focus on helping 50 companies actually deploy something, it will have been worth building. If it becomes another 80-bullet-point presentation that gets filed in a drawer, it will join the long list of well-intentioned European innovation initiatives that confuse activity with outcome.

The catalogue is live. The question is whether anyone will order.