The Mittelstand cooperates on everything else. Why not AI?
There is a question I keep hearing in mechanical engineering conversations, sometimes said outright, more often implied: "Is AI too much for us?" It is almost always the wrong question. Not because AI is simple — it isn't — but because the Mittelstand has been successfully handling things that are too much for any single company for the better part of a century. They just don't call it handling. They call it cooperating.
Look at how German mechanical engineering actually works. You design a machine, but the hydraulics come from a specialist. The control cabinet is built by a Schaltschrankbauer you have worked with for twenty years. The tooling is made by a Zulieferer two towns over. The software integration is done by a Systemhaus. When a new sealing technology arrives that nobody in your building understands yet, you call an engineering office. When you need to certify a new material, you send it to an external testing lab. This is not weakness. This is how the Mittelstand produces quality that the rest of the world cannot replicate. You source the capability you don't have. You keep ownership of the machine.
**The inconsistency**
Against this background, the reflex I see with AI is striking. The same companies that cheerfully call an external Konstruktionsbüro for a design challenge they haven't seen before will describe needing outside help with AI and digitisation as "giving up" or "depending on someone else." The logic is inconsistent. A specialist partner in hydraulics is a sign of good engineering judgment. A specialist partner in AI is somehow a sign of weakness. That doesn't hold.
I think what drives it is partly pride — the Mittelstand earns its identity through deep internal expertise — and partly fear. Fear that AI is so fundamental that outsourcing any part of it means losing control of the company's future. Both feelings are understandable. But the Mittelstand's own history should be reassuring: you have cooperated across company boundaries for decades without losing control of your machines, your quality, or your identity. You kept the direction. You kept the customer relationship. The partner built the part. That is still the deal.
**What the cooperation model actually looks like**
Bringing in an external AI and digitisation partner is not handing over your strategy. It is the same move you make when you call the Systemhaus: you buy the specialist capability you don't have in-house, you stay in the seat, you make the decisions about what matters to your business. A good partner does not replace your engineers. They work alongside them — and they transfer enough to make themselves less necessary over time, not more.
The question worth asking is not "can we handle AI alone?" Most companies cannot, at the pace this is moving, and most do not need to. The question is: which capability do we need to build in-house, which do we source, and who is the right partner for the sourced parts? That is the same question you answer every time you place a Zulieferer contract. You have been answering it well for a long time.
The VDMA Praxistag in Frankfurt raised the strategy question honestly: not how fast can we adopt, but where should this lead us? That question requires internal clarity. The building itself — the data foundation, the agent layer, the integration work — can follow the same cooperative logic that built the machines.