The news
The Eclipse Foundation, one of the largest open-source software organizations, on September 30, 2026 launched the Sovereign AI Foundation, a vendor-neutral body meant to help organizations understand their dependence on AI suppliers and evaluate open-source alternatives. The announcement was made in Brussels.
The new group starts with 17 founding organizations: CEA LIST, EclipseSource, Engineering Ingegneria Informatica, Ericsson, Eurotech, Infosys, Kentyou, KU Leuven, Open Elements, Red Hat, Renesas Electronics, Robert Bosch, Thales, The IO Foundation, TypeFox, the University of York and Vector Informatik.
According to the announcement, the foundation will not develop software or specifications. Instead it will act as a forum for sharing knowledge and assessing technology, connecting the requirements of real deployments with Eclipse projects that already exist and keep their own technical roadmaps.
The projects named in the announcement include Eclipse Theia and Theia AI, for AI-enabled developer tools; Eclipse Enclave, for running AI coding agents in isolated, policy-controlled environments; Eclipse PanEval, for evaluating model capability, safety and cybersecurity; Eclipse LMOS, for building and managing multi-agent systems; and the Open VSX Registry, an extension marketplace the foundation said is approaching 1 billion downloads a month.
Mike Milinkovich, executive director of the Eclipse Foundation, said AI sovereignty is about preserving meaningful choice and control over the systems organizations depend on. Rafee Tarafdar, executive vice president and chief technology officer of Infosys, said in the release that enterprise AI autonomy will be key for organizations to control how they use models.
The Eclipse Foundation said it hosts more than 450 open-source projects and is supported by more than 300 members.
The numbers
- Founding organizations
- 17
- Eclipse Foundation projects
- 450+
- Eclipse Foundation members
- 300+
- Open VSX Registry downloads
- Approaching 1 billion per month
Why CEOs should care
For CIOs and CTOs, sovereignty is increasingly a procurement question: who controls the models, where data is processed and how hard it would be to switch providers. This group gives companies a neutral place to compare notes on those dependencies. A practical first step is an inventory of which AI services your teams rely on and which have no realistic alternative.
For boards and general counsel at companies with European operations or data-residency obligations, the membership list matters. Industrial and telecom names such as Bosch, Thales, Ericsson and Renesas sit alongside software firms and universities, a sign the interest extends to companies that build AI into their products. Ask whether your AI roadmap keeps a viable exit if a vendor changes terms or a regulator restricts data flows.
US vendors selling AI platforms to multinational customers should expect more questions about open-source compatibility, portability and on-premises options. Because the foundation will not ship its own software, its influence will come through guidance and the Eclipse projects it points buyers toward.
The bigger picture
Sovereign AI has become a recurring theme as governments and companies worry about relying on a handful of large AI providers. Most efforts so far have centered on building national models or data centers. The Eclipse approach is different: it bets that open, independently governed tools can give organizations control without building everything themselves. Its success will depend on whether members turn shared assessments into real deployments.
What’s next
Watch for the foundation's first published guidance or assessments, new members joining beyond the initial 17, and whether large US cloud or AI vendors choose to participate.
What “Fact-checked” means
Fact-checking means testing a story’s facts against the evidence before it is published. This story went through at least two separate checks before this version was published.
- What we checked
- Its names, figures, dates, job titles, quotes and who said what were checked against the story’s sources, including its main source where it could be opened. The headline was checked for accuracy and overstatement.
- How
- A first check reviewed the whole story. If it passed, a second, skeptical check went back to the sources to look for mistakes in the most important facts. If a check flagged the story, it was edited to fix the problems found, and a separate re-check then reviewed the whole story again.
- Who
- The checks are made by our newsroom, as steps kept separate from the writing, under rules set by our editor, Hussein Mukhtar. A story the checks still flag is held for the editor, who decides whether it is fixed, published or dropped.
- If something is wrong
- “Fact-checked” does not mean error-free. If a material error is found after publication, we correct the story and add a note saying what changed. Report an error
