Eclipse launches Sovereign AI Foundation to fight AI vendor lock-in

The Eclipse Foundation announced on Wednesday the creation of the Sovereign AI Foundation, a vendor-neutral initiative to help companies manage AI platform lock-in, monitor cyber risk and keep control of their data. The project has been active on Eclipse’s project site since July.

Why switching providers is costly

Mike Milinkovich, executive director of the Eclipse Foundation, told The New Stack that the risk of not taking an open approach depends on how deeply AI is embedded in an organization and how hard the underlying systems are to change. “The deeper the dependency, the greater the potential risk,” he said. “But however embedded the underlying systems are, the costs show up in predictable places: higher prices that make switching difficult, and the challenge of rebuilding integrations, moving data and workloads, documenting systems after the fact, compliance delays, and operational disruption.” Part of the foundation’s work is to compare real-world experiences and build stronger evidence around those risks.

Choice, control and openness

The foundation aims to “preserve choice and control”, promoting open source and open-weight options. Asked whether the initiative carries an America-first message, given that open-weight Chinese models are widely available, Milinkovich said it is not about any single provider. “The same questions apply to AI technologies from the United States, Europe, Asia, or anywhere else,” he said. Access to model weights alone, he added, does not provide the transparency or independent governance organizations may need.

The 17 founding members span technology, industry and academia: CEA LIST, Thales and Kentyou from France; EclipseSource, Open Elements, TypeFox, Vector Informatik and Bosch from Germany; Engineering Ingegneria Informatica and Eurotech from Italy; Ericsson from Sweden; Infosys from India; KU Leuven from Belgium; Renesas Electronics from Japan; The IO Foundation from Malaysia; plus Red Hat and the University of York in the UK.

Where the need is clearest

Milinkovich said the need is most acute in systems with long lifecycles, demanding safety and security requirements, or sensitive data: automotive, manufacturing, telecommunications, aerospace and defense, energy and the public sector. A manufacturer may need AI to keep working with limited connectivity, while a public agency may need to explain how an AI-assisted decision was made.

Early examples include EclipseSource’s work with industrial customers on Eclipse Theia AI, the vendor-neutral Open VSX Registry, Eurotech’s use of Eclipse-governed projects to bring AI to the edge, and Eclipse LMOS, which originated at Deutsche Telekom and now underpins large-scale multi-agent systems. Milinkovich said the common thread is modularity: keeping models, infrastructure and applications separate enough that one can change without rebuilding everything.