Digital Sovereignty and the Crucial Role of Data in AI

As nations like France and Germany push digital sovereignty to the forefront of their policy agendas, a fundamental question is emerging with increasing urgency: who actually owns the data driving Europe's AI systems? With AI models becoming increasingly commoditized, competitive advantage is significantly shifting to the data layer, making data ownership and control strategic elements. In an AI-driven economy, proprietary data—not models themselves—creates opportunities for differentiation and value.

In this context, startups like Countly position themselves as product analytics and customer engagement platforms, built on an open-source, self-hosted foundation. The goal is to help organizations reduce reliance on third-party platforms by enabling them to capture, analyze, and act on their own user data, while maintaining full control over it. The company operates on the belief that data privacy and actionable insights are fundamentally interconnected. Onur Alp Soner, CEO and co-founder of Countly, emphasizes how his company, founded in 2013, anticipated this trend, building its offering around self-hosted analytics long before data sovereignty became a central issue in Europe.

The Strategic Value of Data Ownership in the AI Era

Embedding data sovereignty into a company's commercial strategy can strengthen product differentiation, build regulatory trust, and unlock new partnership opportunities, particularly in sectors where data control is critical, such as healthcare, finance, and public services. Countly's platform, built on an open-source, self-hosted foundation, helps companies collect and process operational and usage data from their software products, essentially enabling them to understand how users interact with apps and services and improve those experiences.

According to Soner, Countly's main focus has always been data control and ownership. The basic idea is simple: if a company doesn't control its data, it doesn't control its systems. This concept, which has guided the company since day one, has become even more relevant with the evolution of the debate. While in the past the issue was primarily linked to data collection practices by third parties for advertising purposes, with the introduction of GDPR and, more recently, with the advent of AI, attention to data ownership has grown exponentially. AI, in particular, is making data ownership economically viable, transforming the data that fuels machine learning systems into a highly valuable company asset, whether directly (selling data) or indirectly (using data to generate revenue). For those evaluating on-premise deployments, understanding these trade-offs is crucial, and resources like those offered by AI-RADAR on /llm-onpremise can provide analytical frameworks for informed decisions.

The Paradox of Sovereignty and Europe's Path Forward

Despite Europe's aspiration for digital sovereignty, there is a clear paradox: many European companies still rely on infrastructure, analytics tools, or models developed by US or Chinese companies. Soner admits that even Countly, while building infrastructure for data sovereignty, still relies on technologies from these origins, as most major databases, for example, are US-based. The question, therefore, is not whether to use external technology, but how to use it, adopting a layered strategy.

Data sovereignty translates into a deliberate architectural strategy: deciding what stays in-house and what can be outsourced. This involves controlling data flows, deciding what information leaves the system, and building proprietary datasets. The goal is to use global tools, but on one's own terms, maintaining control of the data layer, which is the true source of long-term competitive advantage. Companies like Instagram, Amazon, Uber, and Airbnb are, at their core, data businesses. Blindly using tools without thinking about data flows means losing that advantage. For Europe, the challenge is not only to build innovative companies but also to retain them, by creating an attractive ecosystem in terms of funding, incentives, and support—essential elements for keeping talent and businesses within the region.

Building the Future: Data Layer Control as a Strategic Lever

Soner suggests that data ownership should be built into a company's culture early on. Every decision, even for small companies, should go through the lens of data control. It's not just about regulatory compliance, but about recognizing that one's data represents the only truly unique and distinctive aspect of one's business. AI amplifies the quality of the data it is trained on; without control over that data, companies risk outsourcing their long-term competitive advantage.

The AI debate in Europe often focuses on the first two layers of the ecosystem: compute (GPUs, infrastructure) and models (LLMs). However, the data layer, while equally, if not more, important, receives less attention. The real opportunity for Europe lies not in competing directly on models, but in owning and controlling the data layer that underpins them. This means building its own systems, supporting its own companies, and retaining its own talent. Only by doing so, using global technologies on its own terms, can Europe define its own future and capture value in the age of artificial intelligence.