Germany backs Nvidia in its AI hurry

Euronews reported Berlin plans a nine-figure euro contribution to an Nvidia-linked AI project.

Germany has decided that being warned about falling behind in artificial intelligence is apparently less painful than staying behind. At an investment summit in Berlin, Chancellor Friedrich Merz and Nvidia chief executive Jensen Huang agreed on an AI cooperation project that Euronews, citing Handelsblatt, said would include a nine-figure euro contribution from the federal government.

That figure matters less as a trophy than as a confession. Germany is not putting public money into Nvidia because life was going perfectly well without it. It is doing so because the country’s industrial model now faces a problem it understands very well in theory and has moved on slowly in practice: the value in manufacturing is drifting from machines alone toward the data, software and compute wrapped around them.

The reported project is expected to involve about a dozen German companies, including Siemens, Audi and Black Forest Labs, according to Euronews. The stated pitch from Merz and Huang is familiar enough: combine German engineering with cutting-edge AI and preserve the industrial base. That is sensible as far as it goes. Germany still has something many countries would like to rediscover after outsourcing it away — a dense cluster of industrial companies with real processes, real customers and large stores of operational data.

The awkward part is that having the raw material is not the same as using it. Huang’s message in Berlin was not subtle. Euronews reported that he said it pained him to see how far Germany had fallen behind and urged the country to “get to work.” He also repeated an argument he has made before: treat data as a national resource and use it to build what he calls “national intelligence.”

There is strategy in that rhetoric, not just theatre. Nvidia does not merely sell chips; it sells urgency. For governments and manufacturers, the fear is not that AI will replace every factory worker by breakfast, but that competitors with better software, better models and better access to compute will extract more output from the same physical assets. In that contest, industrial data becomes a source of productivity, not just a compliance burden sitting on a server.

Merz used the summit to push German companies to open their industrial data to AI, saying competitiveness in sectors such as autos depends on it, Euronews reported. That is the more consequential point. Subsidising one project is easy enough by state standards. Convincing established companies to share and structure valuable internal data, and to do so across a fragmented industrial landscape, is harder. The denominator here is an economy, not a press release.

Euronews also reported that Merz said the Berlin summit produced investment commitments worth €800 billion through 2028, while Amazon separately announced plans to invest about €16 billion in Germany in 2027, including €4.8 billion focused on AWS around Frankfurt. Big numbers travel well at summits. What they prove is narrower. Commitments are not the same as deployed capital, and deployed capital is not the same as higher productivity.

Still, Germany’s move with Nvidia looks more concrete than the usual “Europe must innovate” sermon. It identifies a bottleneck — access to advanced AI infrastructure tied to industrial use cases — and puts public money behind it. Whether that is a catch-up plan or an expensive reminder of how late the catch-up started is another question. The event in Berlin at least answered one smaller one: Germany heard the insult, and reached for its wallet.