Germany is stepping up efforts to reduce reliance on U.S. big tech in key artificial intelligence (AI) and cloud projects for its military. The shift shows “sovereign AI” priorities spreading into military procurement, putting data control and national security ahead of a simple technology race.

Cryptopolitan reported on April 28 that Thomas Daum, a vice admiral who serves as inspector for cyber and information space in the German armed forces, took a negative view of the possibility of adopting software from U.S. data analytics company Palantir for the military’s next-generation secure cloud project.

The project is a private cloud infrastructure the German armed forces are building to support data processing and AI applications. Germany’s defence ministry sees it as a core foundation for the future digital battlefield.

Daum said the core issue is not technical performance but the operating structure. Germany already uses analysis outputs produced by NATO and some member states through Palantir’s Maven platform, but he raised concerns that external personnel, particularly those linked to Palantir, are involved in operating the system.

He said it is hard to imagine, at least for now, granting a private U.S. company access rights to Germany’s national databases. This effectively means Germany cannot hand control over core military data to a foreign company. Germany has narrowed candidates for the project to three firms: German companies Almato and Orkrist, and French company ChapsVision. The companies are set to go through testing procedures this summer, and the final contract is expected to be signed before year-end. As a result, assessments say it has become more likely that the German military’s AI infrastructure project will be reshaped around a European supply chain rather than U.S. suppliers.

Germany’s wariness is also linked to political factors. German Defence Minister Boris Pistorius has previously voiced concern that Palantir co-founder Peter Thiel held part of the stake in German drone maker Stark Defence. At the time, Germany’s defence ministry approved the related contract after receiving assurances that Thiel would not exercise management or operational control.

The trend is not limited to Germany. The Stanford Institute for Human-Centered Artificial Intelligence said in a recent report that governments are competing to secure “AI sovereignty” to reduce dependence on specific countries and a small number of suppliers. Britain has decided to invest 500 million pounds in a domestic sovereign AI organisation, and France and Brazil are also strengthening home-focused AI regulatory frameworks. China is also cited as a major AI power capable of competing with the United States.

The United States, by contrast, appears to be pushing back against such moves. A U.S. State Department cable made public in February said Secretary of State Marco Rubio instructed diplomats to lobby against overseas data sovereignty legislation. The U.S. government sees national regulations, including the European Union’s General Data Protection Regulation (GDPR), as potentially restricting overseas business by U.S. AI and cloud companies.

Germany’s decision could also have significant repercussions for the U.S. AI industry. Major AI companies such as SpaceX, OpenAI and Anthropic have continued growth strategies while absorbing large-scale investment and losses, and have presented expansion into overseas government markets as one of their key revenue sources.

OpenAI in particular has been reported to be facing internal assessments that user growth and the pace of meeting revenue targets are falling short of expectations. Some board members have reportedly expressed concern about CEO Sam Altman’s strategy of large-scale investment in data centres.

The industry sees that if defence and public infrastructure orders shift away from U.S. companies toward domestic and regional supply chains, adjustments may be unavoidable in the global market growth outlook presented by U.S. AI firms. Analysts say the German military’s choice shows that the benchmark in the AI competition is moving beyond performance toward data sovereignty and national security.