In a world increasingly defined by geopolitical tensions and the overwhelming dominance of American tech giants, Europe’s quest for digital sovereignty has become a critical battleground. At the forefront of this movement is Deutsche Telekom, which is making bold claims about a new generation of “sovereign clouds” designed to break free from foreign dependencies. To unravel the complexities of this strategy, we’re speaking with Vladislav Zaimov, a seasoned specialist in enterprise telecommunications and the risk management of vulnerable networks. We’ll explore the real-world challenges of achieving true technological independence, the strategic tightrope walk of partnering with US firms like Nvidia, and whether a European telecom can realistically challenge the hyperscalers in the age of AI.
Deutsche Telekom’s leadership has outlined achieving 100% data and operational sovereignty but acknowledges a gap in technical sovereignty, relying on partners like Nvidia for chips. How does this reliance impact the overall promise of a “sovereign cloud,” and what are the practical steps to close that technical gap?
It’s a crucial distinction that gets to the heart of the sovereignty debate. When you hear the leadership, like CEO Timotheus Höttges, claim “100%” sovereignty, they’re referring to data and operations. This means your data is stored locally in German data centers, managed by vetted German employees under local laws—a tangible and reassuring promise for many enterprise clients. However, the “technical sovereignty” piece is where the ideal meets reality. Höttges himself is quite candid about this; he’d love to use European or German chipsets, but they simply don’t exist at the required scale or performance level. The reliance on Nvidia isn’t just a preference; it’s a necessity to stay competitive, especially in the AI space. This reliance fundamentally tempers the promise of absolute sovereignty, creating a dependency at the most foundational hardware layer. The practical step, for now, isn’t about replacing Nvidia overnight but about mitigating the risk through what they call “minimizing lock-in” and embracing open technologies. This is the most sovereign solution they feel is achievable today, even if it falls short of the ultimate goal.
The new Industrial AI Cloud in Munich is positioned as being completely free of Huawei products, yet the broader T Cloud Public retains some legacy ties. Can you elaborate on the operational and security challenges of managing this dual approach and how you communicate this distinction to customers?
Managing this dual approach is an incredibly delicate balancing act. On one hand, you have the new, pristine Industrial AI Cloud in Munich, which is presented as a clean break—a spokesperson was emphatic that “No Huawei products will be used at the Munich location.” This is a clear marketing and geopolitical signal. On the other hand, the legacy T Cloud Public, which grew out of the Open Telekom Cloud, was built in partnership with Huawei on OpenStack. Trying to untangle that is a massive undertaking. The primary challenge is maintaining operational and security integrity across two systems with fundamentally different pedigrees. You have to ensure there’s a complete and verifiable separation, which is complex from a network architecture and management perspective. For customers, the communication has to be crystal clear. They need to understand that the “Germany Stack” vision applies to the new AI cloud, while the public cloud still has these historical roots. It’s a message of transition—they’re moving toward greater independence, but the past doesn’t just disappear overnight.
US hyperscalers invest tens of billions annually in R&D, vastly exceeding European telco spending. Given this resource mismatch, what is the specific strategy to achieve “full core feature parity” by 2026, and what unique value proposition can a regional sovereign cloud offer that giants cannot?
The resource gap is staggering and, frankly, a bit terrifying if you’re a European competitor. We’re talking about Deutsche Telekom’s R&D budget of €21 million compared to Microsoft’s nearly $32.5 billion. You can’t outspend them, so you have to outsmart them. The strategy to reach feature parity by 2026 hinges on being laser-focused and leveraging open-source ecosystems like Kubernetes, which was initially developed by Google. This allows them to build on the shoulders of giants without recreating every single component from scratch. The unique value proposition isn’t about having more features or a bigger service catalog. It’s about trust and control, which are becoming paramount in this new geopolitical climate. For a European enterprise, the promise that their data will never be subject to a foreign court order, or that services won’t be suspended due to a transatlantic political spat, is an incredibly powerful driver. This is a value that AWS or Google simply cannot guarantee in the same way, no matter how much they invest in R&D.
The Munich AI facility is a 50/50 venture with Nvidia, including shared investments and revenues. How does this deep integration with a US tech giant align with Europe’s broader digital sovereignty goals, and what specific measures are in place to prevent technological or commercial lock-in?
It’s a pragmatic compromise. On the surface, a 50/50 venture with a US tech behemoth, right down to splitting the €1 billion investment, seems to run counter to the goal of European sovereignty. You’re tying your flagship AI offering directly to an American company. However, the alternative is falling so far behind in AI capabilities that the sovereignty debate becomes moot. The alignment comes from the structure of the deal and the operational control. Deutsche Telekom maintains operational sovereignty, meaning their own people are running the show. To prevent lock-in, the strategy relies heavily on using open-source platforms like Kubernetes and developing their own container-as-a-service platform. This creates an abstraction layer that, in theory, prevents the entire system from being inextricably tied to one vendor’s proprietary hardware or software stack. It’s a calculated risk, betting that they can leverage Nvidia’s world-class technology without becoming a captive customer.
European cloud providers have seen their market share fall from 29% to around 15% in recent years. With US firms now dominating the market, what specific go-to-market strategies and customer incentives are being deployed to reverse this trend and win back enterprise clients?
The numbers paint a bleak picture—a market share cut in half while US hyperscalers have captured about 70% of the European market. To reverse this, the go-to-market strategy has shifted from competing on features to competing on sovereignty and specialized use cases. The “Industrial AI Cloud” is a perfect example. It’s not a generic cloud for everyone; it’s a targeted offering for German industry. The incentive is a combination of fear and opportunity. The fear is geopolitical instability and an unpredictable America; a homegrown cloud feels like a safe harbor. The opportunity is in building a local ecosystem with partners like SAP, creating a “Germany Stack” that is finely tuned to the needs of local businesses. While T-Systems as a whole only grew 2.3% in a recent quarter, the sovereign cloud services segment saw “stunning” 23% growth. This shows that the niche strategy is resonating with a specific, and growing, customer segment that prioritizes data residency and regulatory compliance above all else.
What is your forecast for the European sovereign cloud market over the next five years?
Over the next five years, I forecast that the European sovereign cloud market will become a highly contested and strategically vital niche, even if it doesn’t reclaim the overall market share from hyperscalers. The trend of declining market share for local providers has likely bottomed out, and we’ll see modest but significant growth, driven almost entirely by geopolitical anxiety and stringent regulations. The narrative has fundamentally shifted. It’s no longer just about price and performance; it’s about digital resilience. We will see more specialized clouds emerge, like the Industrial AI Cloud, targeting specific sectors like manufacturing, healthcare, and public administration where data sovereignty isn’t just a preference but a legal necessity. While providers like Deutsche Telekom won’t beat Amazon on scale, they will successfully carve out a defensible and profitable segment of the market by positioning themselves as the trusted local alternative in an increasingly uncertain world.
