AI Sovereignty Starts with Open Infrastructure

Decision makers, industry, and world leaders gathered to discuss the future of artificial intelligence at the event in Paris. Several major investment announcements and policy commitments made it clear that ethical, sustainable, and sovereign AI technologies are strategic priorities for Germany and Europe, and they are willing to invest the necessary resources to successfully compete in this field while holding those important values. When it comes to digital sovereignty, one element was largely missing from many of the conversations: the strategically important open source software infrastructure that makes these AI advancements possible in the first place.

At the Public AI Congress, the Sovereign Tech Agency took part in a panel discussion with many experts from all over Europe, underscoring the importance of open digital infrastructure for the development of sovereign AI. While much of the summit focused on AI applications, we brought a much-needed perspective by shifting the conversation toward the open digital infrastructure that enables competitive development and true innovation in AI. Without a strong and resilient digital foundation to keep AI development accessible and open to competition, Europe risks being dominated by a few powerful players who can dictate the rules of AI access and innovation.

Today’s AI technologies rely on a vast open source ecosystem of programming languages, frameworks, and libraries. The importance and value of openness and enabling competition are reflected in the European Commission's AI strategy for supporting startups and SMEs. From the investments into "AI Factories", which are emphasized as open ecosystems, to the Open Data Directive, and last but not least with the "GenAI4EU" initiative which has a goal of establishing large open innovation ecosystems.

Yet, despite all these substantial investments into creating open AI ecosystems, these investments rarely flow into open digital infrastructure. This leaves much open source software that is critical to the implementation of these plans severely underfunded—and the cracks are starting to show. From security incidents like xz-utils to projects switching to more restrictive licenses, it's clear that this infrastructure is at risk.

If access to these essential technologies becomes restricted due to a lack of investment, Europe’s ability to innovate, compete, and set its own ethical and sustainable standards in AI will be severely compromised. Weak open source infrastructure means that only companies with large security teams and developer resources can fully benefit from these libraries. In fact, the general lack of investment into this strategic resource has allowed a few dominant players to play an outsized role in shaping our open digital infrastructure. This disproportionate control allows them to shape the open source ecosystem to serve their own needs, even going as far as locking out any potential competition and stifling innovation.

This is where the Sovereign Tech Agency plays a crucial role. Recognizing that AI development is only as strong as the digital infrastructure it relies on, we have committed to strengthening the foundational technologies that keep AI accessible, resilient, and democratic. Research shows that 70–90% of all software depends on open source, and AI is no exception—whether open or proprietary.

The impact of AI’s growth on key programming languages and libraries is already visible. Python, for example, is the most popular programming language due to its role in data science and machine learning—it is now twice as popular as the next language on the list. We have commissioned work that has brought important security and stability improvements to Python and PyPI (Python’s packaging system), ensuring developers have reliable access to the libraries they need to build AI applications.

AI innovation also depends on data processing capabilities, which is why we’ve invested in high-performance computing libraries and languages such as OpenBLAS and Fortran. These tools are essential for handling large datasets and performing complex calculations efficiently. Additionally, foundational frameworks like FFmpeg empower developers to analyze and process massive amounts of video and audio data, driving AI applications in media, security, and accessibility. Without these robust tools, the barrier to entry for AI innovation would be prohibitively high.

The Sovereign Tech Agency is committed to securing this foundational infrastructure. By strategically investing in the open source technologies that underpin AI, we ensure that we have the tools to develop, govern, and benefit from AI on our own terms.

Technological sovereignty isn’t just about keeping up with the latest AI advancements—it’s about laying a solid foundation so that investments can drive real and widespread economic progress that aligns with Europe’s priorities. The AI Action Summit made it clear that the stakes today—and the level of commitment—are even higher than before. Let's invest more in digital infrastructure today, so that Europe's capacity for innovation doesn't just keep pace but leads the way.