A guest contribution by Prof. Dr. Dennis-Kenji Kipker
In March 2026, the German federal government finally adopted its long-awaited National Data Centre Strategy, which is firmly anchored in the coalition agreement. The strategy aims to double total data centre capacity by 2030 and quadruple capacity for artificial intelligence and high-performance computing, supported by 28 measures covering energy, land and technological sovereignty.
A high jump from a low starting point
Only through comparison does the ambition of this goal become clear: Germany currently has less than two per cent of global AI computing capacity, despite contributing around four per cent of global gross domestic product. Germany’s total data centre capacity is around three gigawatts, whereas the United States has already reached approximately 48 gigawatts. Even a fourfold increase would therefore do little to narrow this gap, especially as international competitors continue to expand their capacity. The targets themselves are based not on mere estimates, but on model calculations by think tanks and consulting firms. Depending on the level of ambition, these calculations put the required capacity at between 0.8 and 5.9 gigawatts by the end of 2028. This wide range shows how uncertain forecasts about the spread of AI and associated data centre capacity remain.
No competition in Germany for the largest language models
While US technology companies are planning combined investments of around 650 billion dollars this year alone and the People’s Republic of China is investing approximately 295 billion dollars over five years in a state-directed network of AI data centres, Germany's strategy does not seek to replicate a leading model. Such an effort would require capital expenditure that few European companies could afford alone. The focus is therefore on the capacity needed to use AI applications in business and public administration, industry-specific models trained to work with German industry's extensive production and machine data, and sovereign computing capacity for particularly sensitive areas such as critical infrastructure, healthcare, and defence. Although experts estimate that the timeframe for using this wealth of industrial data profitably for the national economy before other countries catch up is only two to four years, it is still regarded as a genuine advantage for Europe as a business location.
Sovereignty remains a continuum
The example of the French AI company Mistral, which expanded its partnership with the US corporation Microsoft again in July 2026, shows how difficult it is to achieve complete digital independence. Although this is officially justified in terms of digital sovereignty, it tends to deepen dependence on a US hyperscaler rather than reduce it. Therefore, at least initially, a graduated approach to sovereignty according to criticality seems more realistic: Europe would primarily retain control where data protection, availability and security matter most, while continuing to rely on global providers in less sensitive areas for the time being. Further deep-seated technological dependencies will probably only be reduced over the coming years, as they were built up more or less deliberately over the preceding decades.
Consequently, the federal government’s new Data Centre Strategy provides a coherent framework with concrete figures and deadlines for the first time. However, this will not automatically enable Germany to catch up with the world’s IT leaders. Whether it becomes more than a well-intentioned political plan will depend, above all, on how quickly permits are granted, land is designated, and energy prices are stabilised.
As a result, the federal government’s new Data Centre Strategy provides a coherent framework with precise figures and deadlines for the first time. However, this will not automatically enable Germany to catch up with the world’s IT industry leaders. Whether it becomes more than a well-intentioned political plan will depend above all on the speed with which permits are granted, land is designated and energy prices become competitive, without disregarding environmental and social concerns that are currently the subject of intense domestic debate in the United States and are already causing considerable upheaval there despite all the innovation.
