OpenAIRE has showcased a series of AI-driven research innovations at the Open Science Conference 2025, highlighting new tools that advance transparency, reproducibility, and evidence-based analysis across the European research ecosystem. Through posters and live demonstrations, the organization presented practical applications of AI designed to enhance transparency, accessibility, reproducibility, and reusability, which are core principles of Open Science.
One of the featured contributions was a poster within the EOSC Track project titled Enhancing Discovery, Policy, and Practice: A Hands-On Demo of the European Open Science Resources Registry. The presentation introduced the newly developed European Open Science Resources Registry, an AI-enhanced component of the EOSC Open Science Observatory. This interactive registry curates key Open Science documents—including policies, strategies, and best practices—and applies natural language processing and machine learning to support intelligent search, classification, and evidence-based policy evaluation.
Insights shared during the session compared the registry’s monitoring capabilities to a system that provides clear indicators of strengths and areas for improvement, similar to how a wearable device tracks personal metrics. This framing emphasized how AI can help monitor Open Science policies and practices across Europe and strengthen discovery workflows. The session also allowed attendees to explore the EOSC Open Science Observatory, examine the new AI-powered Resources Registry, and discuss OpenAIRE’s contributions to evidence-based Open Science monitoring with stakeholders shaping future policy and infrastructure. The poster is available on Zenodo.
A second live demonstration, Powering Open Science Across Borders: A Live Demonstration of the EOSC EU Node, showcased how the EOSC EU Node supports researchers while aligning with Open Science principles. The demo focused on practical usability, emphasizing how hands-on sessions help communities better understand the value and impact of European Open Science services.
A third contribution, presented through the poster The Intelligence Behind the OpenAIRE Graph: Linking Science with AI, demonstrated how the OpenAIRE Graph integrates diverse research sources into a richly linked, machine-actionable ecosystem. The AI-driven analytical workflow strengthens data quality, connectivity, and usability. Insights highlighted the importance of transparency and explainability for building trust in AI-driven workflows, as well as the need for training and capacity building to enable responsible use of AI by researchers.
Discussions throughout the conference reinforced several key observations about the future of Open Science infrastructure. Transparent and explainable AI integration is essential for maintaining research integrity and community trust. Training and capacity building are increasingly critical to ensure researchers across regions and career stages can effectively employ AI tools while adhering to Open Science values.
Additionally, as providers of foundational research infrastructure, organizations must uphold data sovereignty and preserve the control of European research communities over their scholarly outputs and services. The convergence of AI and Open Science was recognized as an opportunity to build more resilient, transparent, and impactful research ecosystems for the public good.
Feedback gathered during the event will inform the ongoing development of the OpenAIRE Graph and guide future enhancements aligned with community needs. Across OpenAIRE’s participation, several shared insights emerged: AI serves as both an enabler and a challenge for Open Science and must be applied transparently and ethically; FAIR Digital Objects and linked data are becoming increasingly important to ensure research outputs are discoverable and machine-actionable; there is a growing demand for training to help researchers use AI responsibly; and collaboration across infrastructures, policies, and communities remains essential to sustaining a trusted, connected research environment.
Click here to read the original press release.
