|

GITEX AI Europe 2026: What the Berlin Agenda Reveals About Europe’s AI Infrastructure Race

ByFluz

GITEX AI Europe returns to Messe Berlin on June 30 and July 1. Last year’s inaugural edition drew 1,400 enterprises and startups from over 100 countries. The 2026 edition scales to 800 companies and startups, 500 investors, 120 speakers. By those numbers alone it qualifies as a significant event. The more interesting question is what the agenda — its themes, its tracks, its structural choices — reveals about where European enterprise AI actually stands in mid-2026.

Events of this scale are not just gatherings. They are a form of collective diagnostic. The topics that earn dedicated forums, the problems still worth a roundtable, the distinctions that have become commercially urgent: read carefully, they tell you more about the state of the market than most analyst reports. GITEX AI Europe 2026 tells a specific story. Europe’s AI conversation has moved from the software layer to the infrastructure layer, and from the question of adoption to the question of control.

Why the second edition matters more than the first

The inaugural GITEX AI Europe, held in Berlin in 2025, arrived at a moment when AI was still largely aspirational for most European enterprises. The agenda reflected that: demonstrations of capability, early use cases, frameworks for thinking about governance. Useful, but still in the mode of building a market.

The 2026 edition arrives after a year in which that market moved. AI is deployed at scale in a majority of large European organisations. According to data compiled for the GITEX AI EUROPE and LUE whitepaper published in May 2026, Europe’s ICT sector is now valued at over €1 trillion. The EU’s InvestAI programme is committing €200 billion to funding five AI gigafactories across the continent, each equipped with 100,000 or more specialised GPUs. Germany alone may need to triple its data centre capacity by the end of the decade, requiring up to €60 billion in investment to meet projected AI workloads.

These are not strategic ambitions. They are procurement decisions already in motion. The second edition of GITEX AI Europe is where the organisations making those decisions, and the vendors serving them, come to work through what comes next. That is a fundamentally different conversation from 2025.

The sovereignty question has moved to inference

The easiest version of the digital sovereignty debate is about data residency: where is the data stored, under whose legal jurisdiction. That version of the debate is largely settled, or at least well understood. The harder version, the one driving the GITEX AI Europe agenda in 2026, is about inference: where do AI models actually run.

Dr. Andreas Nauerz, chief product officer at IONOS, one of Europe’s largest cloud and hosting providers, put the distinction plainly in analysis ahead of the event: “If AI inference consolidates on non-European infrastructure before sovereign compute scales, enterprise AI strategies become jurisdictionally compromised — regardless of where the data lives. That window is closing faster than most cloud roadmaps acknowledge.”

The concern is structural. A European company that trains its AI models on European data, inside a European cloud environment, but runs inference on a US hyperscaler’s infrastructure is still subject to the legal frameworks of that hyperscaler’s headquarters. The US CLOUD Act, extraterritorial jurisdiction mechanisms, and shifting regulatory environments mean that “stored in Europe” and “controlled by Europeans” are not the same statement. For organisations in regulated sectors — financial services under DORA, healthcare under NIS-2, public sector under the EU Data Governance Act — this distinction has moved from theoretical to operationally material.

The GITEX AI Europe whitepaper identifies four pillars shaping Europe’s response: compute, cloud, open source, and capital. The compute pillar is where the urgency concentrates. European data centre capacity is projected to grow by 70 percent by 2030, but AI demand is expected to outpace that growth. Closing the gap requires not just investment, but a shift in who owns the infrastructure being built.

What the agenda tracks actually signal

Conference tracks are editorial choices. They reflect what the organising body believes the market needs to talk about and what topics have commercial density — enough buyers, sellers, and practitioners with real skin in the game to fill the room. The GITEX AI Europe 2026 track structure is worth reading in that light.

The AI Stack Sovereignty Forum, running on June 30, addresses the question of control over the full technology stack: from the cloud layer to the model layer to the inference layer. The AI Factories and Compute Corridors track on July 1 moves the conversation from principle to physical infrastructure — the actual GPU clusters, energy requirements, and network architecture that large-scale AI production requires. Scaling Enterprise AI spans both days, with a focus on what it takes to move from pilot to production inside complex organisations.

The cybersecurity dimension runs through the entire event via GISEC Europe, the co-located forum for security professionals. Richard Werner, cybersecurity platform lead for Europe at TrendAI, observed ahead of the event that hybrid cloud environments — the architecture most European enterprises have settled into — create their own distinct attack surfaces. Complexity introduced by multi-cloud arrangements often produces the gaps that are hardest to close: inconsistent identity controls, fragmented monitoring, misconfigurations that exist precisely because no single team has visibility across the whole stack.

The co-located North Star Europe, featuring 750 startups and 600 investors, rounds out the picture. The startups presenting in Berlin in 2026 have survived a significant funding correction. The valuations and business models on display will reflect what European enterprise buyers are actually willing to pay for, not what was fundable in 2021. That selection effect makes the startup floor one of the more honest maps of where AI investment has landed.

The €180 million sovereign cloud and what it means for buyers

In April 2026, the European Commission awarded a €180 million sovereign cloud tender, the clearest signal yet that digital sovereignty has moved from policy language to procurement decision. The contract structures European cloud infrastructure that guarantees legal, operational, and data control within EU jurisdictions — a direct response to the governance gaps that GDPR alone cannot close.

For enterprise AI buyers, this changes the evaluation criteria for cloud contracts in ways that have not fully propagated through procurement processes yet. Three implications stand out.

First, provider jurisdiction now belongs in the due diligence checklist alongside data residency. The fact that a provider operates a data centre in Frankfurt does not establish EU-controlled infrastructure if the parent company is headquartered in a jurisdiction with extraterritorial data access provisions. Sovereign-first procurement means asking not just where the data sits, but who can compel access to it.

Second, open standards matter more than they did. The Sovereign Cloud Stack, developed with funding from the German Federal Ministry for Economic Affairs, enables organisations to migrate workloads freely between providers. Vendor lock-in and sovereignty are in tension: a provider that offers genuine sovereignty on proprietary infrastructure is offering something that looks like sovereignty but functions like dependence. Open-source foundations reduce that risk.

Third, the timeline is compressing. According to the European Commission’s own projections, 91 percent of enterprise workloads will be in cloud by 2028. Organisations that defer sovereign infrastructure decisions until 2027 will be rebuilding AI stacks already in production, at significantly higher cost and disruption. The window for sequencing sovereignty into the architecture from the outset, rather than retrofitting it, is not indefinitely open.

Three questions to bring to Berlin

GITEX AI Europe runs two days. For organisations attending, the value that travels home depends on the specificity of the questions brought in.

The first is structural: does your AI roadmap depend on inference running on non-European infrastructure? If the answer is yes, the follow-on question is not whether to address it, but when. The €180 million tender, the InvestAI gigafactory programme, and the sovereign cloud standards work are all pointing toward a European infrastructure layer that will be qualitatively different in 2027 and 2028 than it is today. Organisations that have mapped their dependencies are better positioned to time those transitions deliberately.

The second is competitive: where are the European vendors closing the gap? The IONOS example — a provider explicitly building on EU-controlled infrastructure as a differentiator — is one instance of a broader pattern. At GITEX AI Europe, the vendor presentations worth attending are not the ones describing the problem. They are the ones demonstrating production deployments of sovereign AI stack components in regulated industry contexts, specifically financial services, healthcare, and industrial manufacturing.

The third is regulatory: how does your sovereign compute strategy intersect with EU AI Act compliance? The Act’s GPAI enforcement provisions take effect on August 2, 2026. For organisations deploying general-purpose AI models in production, the governance requirements around model documentation, capability evaluation, and systemic risk assessment apply regardless of which cloud layer the inference runs on. But the combination of sovereign infrastructure and EU AI Act compliance creates a coherent framework that non-sovereign deployments make harder to achieve.

For Italian organisations specifically, the PNRR Transizione 4.0 investments in digitalisation and the operational context of public sector AI procurement give these questions a particular urgency. The decisions being made in Berlin in late June will shape European AI infrastructure for years. Arriving with the right questions is half the work.

Fulvio Marchetti is the founder of OneSynergy, a consulting network focused on AI strategy, digital transformation, and innovation management based in Turin, Italy. For more on EU AI Act compliance and governance strategy, see the analysis published in April 2026.

Data in this article reflects publicly available sources as of June 8, 2026.

OneSynergy works with organisations in Turin, Italy and across Europe on technology scouting and the build-versus-buy decisions behind AI and infrastructure adoption. See how we approach digital transformation and technology integration.

Sources

Relevant OneSynergy service

EU AI Act Readiness and AI GovernanceExplore →

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *