Accelerating Open Source AI Adoption: the Inria Foundation Launches New Working Group

13 octobre 2026

How are businesses and public sector organizations adopting open AI models? The Inria Foundation is bringing together AI users and developers to document use cases, identify obstacles, and establish best practices. The initiative is supported by NVIDIA, and its findings will be released in spring 2027.

Adoption, the bottleneck for Open Source

France is a strong advocate for open source AI in Europe, in the name of innovation, technological autonomy, security, and industrial competitiveness. But opening up a model is not enough: its success depends on having real, engaged users. This shift to actual use, which remains poorly documented, is what the working group aims to shed light on.

An open group led by the Inria Foundation

The Working Group on Open Source AI Adoption brings together major industrial users, public sector organizations and AI developers. It has three missions:

  • Document concrete use cases
  • Identify common challenges
  • Establish best practices to accelerate adoption.

NVIDIA, a leader in AI and accelerated computing, is among the initiative’s main contributors and founding members. Companies and organizations such as Bull, Cap Digital, the Caisse des Dépôts-EuroCommons, Cigref, Dassault Systèmes, Crédit Mutuel Alliance Fédérale, through its technology subsidiary Euro-Information, Gradium, H Company, Hugging Face and Scaleway have already joined.

Building on the Foundation’s commitment to open source

Inria and the Inria Foundation have long supported open source software and digital commons, notably through projects hosted by the Foundation, such as Software Heritage and the OCaml community.

Want to contribute to this work?

Does your organization use or develop open AI models and want to share its experience?

Contact Simon Chignard: missionIAOpenSource@fondation-inria.fr

Learn more

Press Contact: Selena Doukhan – selena.doukhan@fondation-inria.fr

Beyond the development dynamic, adoption is the main bottleneck for any open source strategy. It is the existence of a community of real, engaged users that marks a project’s success. […] The explosion of AI is radically reshaping these challenges, particularly when it comes to sovereignty.

Understanding how organizations adopt open source AI, identifying the barriers, finding the levers, and launching mutual support processes: this is how we give Europe the means to make openness a true driver of autonomy.

Documenting concrete use cases and sharing expertise are essential to accelerating the adoption of open source AI at scale and making it accessible across all sectors.

At Bull, we see open source as a key driver of innovation, transparency, and digital sovereignty—essential foundations for building trust in AI and accelerating its long-term adoption, particularly in Europe. Our commitment is reflected both in our direct contributions to open-source projects and in the development of open platforms such as BullSequana AI, designed to prevent vendor lock-in and give our customers the freedom to choose their models, infrastructure, and partners.

We are pleased to be working together to address the challenges of building trustworthy, high- performing, and diverse AI. Cap Digital’s members are deeply committed to freedom of choice, and open-source AI can play an important role in making that possible. We have made it a key priority in our 2027–2030 roadmap. We will ensure that our members have the opportunity to join this emerging community and make meaningful contributions to its work, with rigor and ambition.

Cigref is delighted to join this Fondation Inria initiative. Scaling up open-source AI is a central concern for our members, both in terms of technological control and economic performance. This initiative fits squarely with several of our strategic priorities for the 2026-2027 work cycle and will complement them very usefully.

Dassault Systèmes is building trustworthy AI for industry, grounded in science and designed to empower human ingenuity. This industrial AI relies on Industry World Models that bring together industrial representations, knowledge, and domain-specific expertise. Within this multi-model strategy, horizontal openness is one of the pathways to innovation. Our goal is to give industrial companies greater choice and control over how they use AI, enabling them to create sustainable value.

Open source adoption across sectors enables full control, reliability, competitive progress, as exemplified across the Hugging Face community. We are glad to support this initiative and report

Open-source AI is a major asset for Europe’s technological independence and competitiveness. We are delighted to contribute to the work of the Fondation Inria, with the support of NVIDIA, to help European companies make the most of this opportunity.