The EOSC-ARENA (AI Research Enhancement through Networked Agents) consortium is pleased to announce the official launch of this three-year European Union-funded project on the 1st of June 2026.  

About the project

EOSC-ARENA sets out to deliver a sovereign, generative and agentic Artificial Intelligence (AI) environment integrated with the European Open Science Cloud (EOSC). This AI environment will serve as a scientific assistant supporting the full research lifecycle, from literature review and hypothesis generation to analysis, reporting, and provenance capture. The project aims to address the pressing needs in the use of smart algorithms and AI/ML services in scientific research, and focuses on model and data sovereignty emphasizing human-centred AI, responsible usage, research integrity, transparency, trustworthiness, quality, and provenance.

The project will build an advanced, scalable, multi-agent system and a marketplace for Generative AI (GenAI) agents and services. It will provide federated training and inference, secure generation with augmented search and integrations based on the Model Context Protocol. The EOSC-ARENA results will become part of the AI4EOSC ecosystem and collaboration, extending the  AI4OS software stack towards GenAI. The EOSC-ARENA system will be deployed on EU e-infrastructures and interoperable with EOSC EU nodes. Twelve real-life use cases from different scientific domains are selected to co-design, implement, and assess the effectiveness of the project solutions.  

The EOSC-ARENA vision extends beyond traditional GenAI applications, by working on three different foundation areas: (i) A sophisticated GenAI agentic system, with a strong focus on scientific robustness; (ii) a community building and sound training program focused on responsible usage, enabling researchers to perform informed decisions, aligned with EU values from an ethical and human-centric perspective; and (iii) a comprehensive set of GenAI infrastructure services. All together, these three areas serve as enablers in order to deliver the final sovereign AI ecosystem to enable open science and collaborative Human-AI work.  

The main outcome of the project will see an EOSC-ready platform release with agent execution and marketplace, open-source components, machine-actionable APIs and provenance mechanisms. Of equal importance are policy guidance and training assets, with the aim of accelerating trustworthy AI adoption. The project targets demonstrators integrated with EOSC services and contributes directly to the EOSC and its strategic research and innovation agenda by strengthening interoperability, FAIRness and sustainability of AI in European research.