Open Source and the AI Act: A Developer’s Perspective
GitHub CEO Thomas Dohmke recently addressed European policymakers in Brussels, offering a developer’s view on the upcoming EU AI Act. His core argument: the legislation must be structured to protect the open source community, which he sees as the foundation of Europe’s future AI leadership. The stakes are high, as the AI Act is poised to become a global benchmark for AI regulation.
From Skepticism to Ubiquity: The Open Source Journey
Dohmke drew on three decades of open source history to frame his concerns. In the late 1990s and 2000s, open source was met with widespread skepticism. Thirty years later, that model has proven superior. Open source now powers governments and corporations alike, with collaboration models that Dohmke argues align closely with European values—deliberation, cross-border contributions, and diverse perspectives. GitHub has been active in European policy debates before, including the Copyright Directive and the DSA, and now monitors proposals for CE marking for software.
The AI Wave: A Double-Edged Sword
The current moment is defined by a second wave of innovation: artificial intelligence. Already pervasive in everyday life—from photo sorting to driving—AI is having a transformative impact on software development itself. Tools like GitHub Copilot allow developers to write code in half the time. Crucially, this new era is built on the convergence of AI and open source. Frameworks such as PyTorch and TensorFlow underpin most AI development, while open source models themselves are now shared openly. The open source community will remain critical to advancing AI, tackling challenges like bias testing, security, and transparency.
European Innovation in Practice
Dohmke highlighted stable diffusion as a key example of European open source AI success. The text-to-image model, released last August by researchers at Ludwig Maximilian University in Munich, was trained on thousands of GPUs at a cost of over €500,000. Through open source collaboration, developers achieved significant efficiencies: the model can now run on a smartphone, where it once required high-powered devices. This story, Dohmke argues, helps break the narrative that only the US and China can lead in tech innovation.
The economic stakes are equally significant. A 2021 OpenForum Europe study found that open source contributes up to €95 billion to the EU’s GDP—comparable to the air and water transport industries combined. With AI projected to add more than $15 trillion to the global economy by 2030, the potential for European developers is immense. GitHub is already tracking the economic impact of generative AI.
The Case for Exemption
Dohmke’s central policy ask is simple: open source developers should be exempt from the AI Act. He argues that compliance burdens should fall on entities shipping products, not on a community of individuals—many of whom are volunteers, students, academics, or professionals with other day jobs. They don’t stand to profit from their contributions and lack the budgets or compliance departments of commercial entities.
The AI Act has the opportunity to foster democratized innovation and codify Europe’s leadership in open, values-based AI. If the open source community is given the clarity and space to grow, Dohmke contends, the EU can position itself as a leader in this new age—rather than conceding the field to the US and China. As he put it: the greatest innovations come from the roots, but only if they are given the daylight to grow.
