Learn about EGI’s role in enabling Artificial Intelligence for scientific research and innovation
EGI for AI
About AI in Science
Artificial Intelligence (AI) is transforming scientific research by enabling new approaches to data analysis, modelling, simulation, and knowledge discovery. From machine learning and computer vision to large language models and AI-powered digital twins, AI technologies are increasingly embedded in scientific workflows across disciplines.
Across Europe, numerous initiatives and projects are advancing AI-enabled science, aiming to improve research efficiency, scalability, and reproducibility while addressing challenges related to access to compute resources, data availability, skills, and responsible use. EGI actively supports these efforts by providing federated infrastructure, platforms, and expertise that allow AI to be applied in a trusted and sustainable way.
EGI Support for AI in Science
The EGI Federation actively supports AI in Science through:
Technological Solutions
EGI provides federated infrastructure and services that support the full AI lifecycle, from development and model training to deployment, reuse and operation. These solutions enable access to compute, integrate AI workflows with data, and deploy AI-ready platforms and Digital Twin frameworks in a secure and trusted European environment.
Policy and Governance Support
EGI contributes to governance, trust, and sustainability frameworks that enable responsible and transparent use of AI in research. Its activities support alignment with European policy initiatives and promote openness, FAIR principles, and interoperability across infrastructures. Through its participation in European projects, EGI helps transition AI solutions from pilot activities into sustainable operational services, including contributions to the European Strategy for AI in Science and the Resource for AI Science in Europe (RAISE) through the SCIANCE project.
AI Landscape Harmonisation
EGI helps create synergies between AI-related projects and initiatives by identifying common needs, promoting reuse of platforms and services, and avoiding fragmentation and duplication of effort.
By acting as a connector between research communities, digital infrastructures, and policy initiatives, EGI contributes to a more coherent and efficient European AI in Science landscape.
Contribution to Projects
EGI leads or contributes to several European projects that advance AI adoption in science. These projects allow EGI to co-design solutions with research communities, validate them in real use cases, and evolve successful results into sustainable services.
Focus Areas
EGI’s activities for AI in Science focus on:
Enabling scalable and federated access to AI-ready compute and data
Supporting AI adoption by research communities through co-design and user support
Promoting reuse of AI models, workflows, and and Digital Twin frameworks across disciplines
Aligning AI developments with EOSC, Data Spaces, and European policy initiatives
Contributing to sustainable governance and long-term operation of AI services
Technological Solutions
EGI services support AI in Science by providing integrated capabilities that combine compute, data, and collaborative environments. These services enable researchers to develop, train, deploy, and operate AI models while supporting reproducible workflows, cross-domain collaboration, and reuse of models, tools, and Digital Twin frameworks.
Examples include:
- Federated access to scalable cloud, HTC, and HPC resources, including GPU-enabled environments
- Interactive notebook environments supporting development and reproducible research workflow
- Data Exploitation Platforms (DEPs) integrating AI frameworks with scientific datasets and compute resource
- AI model deployment and serving for operational inference workflows
• Digital Twin frameworks enabling interoperable simulation and data-driven modelling across domains - Federated identity and access management enabling secure and trusted access to distributed resources.
Our Projects and Initiatives
EGI is involved in a broad portfolio of ongoing and past projects that address AI in Science from different perspectives, ranging from platform development to community adoption and coordination.
- Projects
- Use Cases
Ongoing Projects
SCIANCE supports the development of the Resource for AI Science in Europe (RAISE) by coordinating a bottom-up, community-driven approach to AI in Science. The project pilots the RAISE Secretariat and Digital Hub and contributes to shaping a Strategic Research and Innovation Agenda (SRIA) for AI in Science.
RI-SCALE develops Data Exploitation Platforms that integrate scalable AI frameworks with scientific data and compute resources, enabling advanced data analysis and AI-driven insights for multiple research infrastructures.
(coordinated project)
The project explores AI-based services for improved data discovery, dataset linking, and workflow composition within EOSC, supporting more machine-actionable and reusable research data.
(coordinated project)
EUCAIM builds a federated European infrastructure for cancer images and provides experimentation platforms for AI tools supporting precision medicine.
PHENET applies AI-based sensing, predictive modelling, and digital twin technologies to support agroecological transition using multi-source data.
ANERIS develops AI-enabled marine life sensing and monitoring systems for real-time biodiversity observation.
SAGE focuses on environmental data integration and analysis to support the Green Deal Data Space, with potential for future AI-enabled services.
Past Projects
EGI has also gained extensive experience from past AI-related projects, including:
Focused on AI-based image analysis for aquatic environments, highlighting the importance of reusable models, strong user support, and efficient access to GPUs.
(coordinated project)
Developed AI-driven digital twin technologies and workflows, revealing common AI requirements across scientific domains and challenges in integrating AI with HPC environments.
(coordinated project)
Contributed to European AI-on-Demand (AIoD) platforms for sharing datasets, models, and tools, providing valuable lessons on sustainability, governance, and community adoption of AI services.
