From Polish data-centre capacity to production-scale AI
Beyond.pl’s selected announcements centre on expanding sovereign AI infrastructure at its data-centre campus in Poznan. The provider says its AI Factory, launched in September 2025 and built on an NVIDIA reference architecture, offers GPU as a Service, AI as a Service, GPU colocation and managed services for developing, training and deploying AI workloads. It plans to bring 14 MW of AI infrastructure online in 2027 and make another 56 MW of campus power capacity available from 2028, within a campus designed to scale to 100 MW. Beyond.pl presents the service as a path from proofs of concept to production and reports recognition from the 2026 Everpure AI Accelerator Awards and the ITwiz Tech Sovereignty Award category. These awards substantiate external recognition, but the broader capability and sovereignty statements remain provider claims.
Sovereignty is becoming a commercial and industrial strategy
The recurring theme is not simply access to GPUs, but the combination of local infrastructure, European regulatory alignment, data control and scalable computing capacity. This suggests that Beyond.pl wants to occupy a strategic layer between conventional colocation and large-scale AI platforms, serving enterprises, public institutions, researchers and startups that prefer workloads and data to remain under European jurisdiction. Its consideration of the EU AI Gigafactories tender extends that ambition from a private AI Factory towards participation in Europe’s industrial AI policy. At the same time, the reliance on NVIDIA architecture and relationships with infrastructure technology partners shows that Beyond.pl’s version of technological sovereignty is primarily about operational control, location and governance rather than independence from non-European hardware ecosystems.
Expansion depends on capital, partnerships and demonstrable demand
The scale of the plans creates material execution risks. In September 2026, Beyond.pl said it had not decided whether to bid for an EU AI Gigafactory and was still reviewing legal, financing and partnership models. The provider cited an indicative investment range of EUR 3–5 billion for one such project under EU assumptions, making external finance and consortium formation central to any bid. Its own capacity roadmap is also forward-looking: the articles do not provide utilisation levels, committed customer demand, pricing, energy sourcing, delivery milestones or independently tested workload performance. Nor do the award announcements demonstrate that planned capacity will be delivered or commercially absorbed. Buyers should therefore separate infrastructure already available from capacity scheduled for 2027 and 2028, and assess how contractual service levels, supply-chain dependencies, security controls and data-governance arrangements apply to their workloads.
Decision guide: Beyond.pl’s AI Factory expansion plan
For IT decision-makers in Eastern Europe, Beyond.pl is relevant because it is attempting to establish Poland as a regional location for high-density AI computing rather than merely a destination for standard data-centre services. Its local campus, production-oriented GPU services and emphasis on European data control may merit evaluation for regulated or latency-sensitive AI deployments. However, procurement decisions should be based on currently contracted capacity and verifiable operational terms, not solely on awards, sovereignty language or future megawatt targets. Buyers should compare the provider’s account of its AI Factory roadmap and Everpure recognition, its assessment of the EU AI Gigafactories tender and its ITwiz sovereignty award result, then request evidence on availability, performance, compliance, resilience and expansion funding. The strategic direction is clear: Beyond.pl is betting that sovereign, locally operated AI capacity will become an important part of Central and Eastern Europe’s cloud landscape. Whether it becomes a regional-scale platform will depend on execution and sustained customer demand.