Europeana and Open Future Release Joint Impulse Paper on AI-Era Data Publishing
Europeana Foundation, in partnership with the Open Future Foundation, has published a new Impulse Paper titled "Publishing cultural heritage data in the age of AI." Released in early December 2025, the document provides cultural heritage institutions across Europe with a practical framework for deciding how and when to make their collections available as training data for artificial intelligence systems. The paper focuses particularly on generative AI and its implications for data sharing in the sector.
Context Within Europe's Common European Data Space for Cultural Heritage
The Impulse Paper forms part of ongoing work on the common European data space for cultural heritage, an EU initiative under the Digital Europe Programme. This data space aims to support the digital transformation of Europe's cultural sector by enabling institutions to share digitised content with high-quality metadata, including in 3D formats. Europeana serves as a central platform in this ecosystem, aggregating content from thousands of institutions and facilitating reuse in research, education, and creative industries.
European universities and their libraries, archives, and museums play a central role in this landscape. Many university collections contribute to Europeana, and academic researchers in digital humanities increasingly rely on large-scale, machine-readable datasets for computational analysis. The new paper addresses growing questions about how institutions should handle requests for data to train AI models while protecting public-interest values.
Key Challenges Addressed in the Impulse Paper
Cultural heritage institutions face complex decisions when considering AI training uses. Generative AI systems require vast amounts of data, raising questions about consent, attribution, commercial exploitation, and the potential for outputs that misrepresent or decontextualise heritage materials. The paper highlights both opportunities—such as improved discoverability, new research insights, and innovative educational tools—and risks, including loss of control over how content is used and potential conflicts with existing licensing frameworks.
For higher education institutions, these challenges intersect with open science principles, research integrity, and compliance with the EU AI Act. University libraries and digital humanities departments often manage collections that straddle public domain materials and copyrighted works, making clear guidance on tiered access particularly relevant.
The Proposed Tiered Access Framework
The Impulse Paper advocates a tiered model that distinguishes between human and machine uses of cultural heritage data. Under this approach, institutions can offer different levels of access: fully open for human viewing and research, more controlled or conditional for machine learning and AI training purposes, and restricted where commercial or high-risk uses are involved. The framework is grounded in public-interest considerations, encouraging institutions to weigh factors such as the nature of the content, the intended use, and the potential benefits to society.
Authors, including Paul Keller with contributions from Alek Tarkowski, Antoine Isaac, Ariadna Matas Casadevall, Harry Verwayen, and Lorena Dianel Aldana, present the model as an invitation for dialogue rather than a rigid prescription. It aligns with broader European efforts to build trustworthy AI ecosystems while preserving the cultural sector's stewardship role.
Photo by Julia Taubitz on Unsplash
Relevance for European Universities and Academic Research
University collections and research infrastructures stand to benefit significantly from clearer guidance on data publishing. Digital humanities scholars at institutions across Europe use Europeana APIs and datasets for projects in history, literature, archaeology, and linguistics. The Collections as Data movement, supported by Europeana and partners such as DARIAH (Digital Research Infrastructure for the Arts and Humanities), promotes the release of machine-readable datasets specifically for computational research.
The Impulse Paper's recommendations can help university libraries develop policies that support both open research and responsible AI development. Webinars and training initiatives by Europeana on data reuse for higher education and research already demonstrate strong demand from academic communities. By adopting tiered approaches, universities can facilitate AI-driven discoveries while maintaining ethical oversight.
Connections to the EU AI Act and Broader Policy Landscape
The timing of the Impulse Paper coincides with implementation of the EU AI Act, which introduces risk-based rules for AI systems. Cultural heritage data used in high-risk applications may require additional safeguards. The paper's public-interest framework offers institutions a way to navigate these requirements proactively.
European higher education institutions involved in AI research and digital heritage projects are encouraged to engage with the Alignment Assembly on Culture for AI, the collective intelligence process that informed the Impulse Paper. This participatory approach reflects the sector's collaborative tradition and supports the development of shared standards.
Opportunities for Innovation in Teaching and Research
Clearer data publishing guidelines can accelerate innovation in European universities. AI tools trained on responsibly shared cultural heritage data can enhance virtual exhibitions, personalised learning platforms, and advanced text and image analysis in the classroom. University researchers can explore new questions about European history and culture at scale, while students gain hands-on experience with ethical data practices.
Examples from the sector include workflows developed in collaboration with DARIAH for publishing Collections as Data and ongoing work on the European Open Science Cloud. These initiatives demonstrate how the Impulse Paper's principles can be put into practice within academic settings.
Stakeholder Perspectives and Next Steps
Leaders from Europeana and Open Future emphasise that the paper is intended to spark conversation among cultural heritage professionals, policymakers, technologists, and researchers. University representatives are particularly encouraged to participate, given their dual role as content providers and heavy users of digital collections.
Future activities may include workshops, pilot projects, and updates to Europeana's existing publishing and licensing frameworks. Institutions are invited to test the tiered model with their own collections and share feedback to refine the approach.
Implications for European Higher Education Policy
The Impulse Paper contributes to wider discussions on research data management, open access, and digital skills in European universities. Ministries of education and research funding bodies, along with organisations such as the European University Association, are likely to monitor developments as they shape institutional policies and funding calls under Horizon Europe and the Digital Europe Programme.
By supporting responsible data sharing, the framework can help universities meet performance indicators related to research impact, international collaboration, and innovation. It also aligns with efforts to strengthen Europe's technological sovereignty in AI while upholding cultural values.
Future Outlook for Cultural Heritage and AI in Academia
As generative AI continues to evolve, the principles outlined in the Impulse Paper will become increasingly important for European higher education. Universities that adopt thoughtful data publishing strategies stand to lead in ethical AI research and digital heritage studies. The paper provides a timely foundation for balancing openness with responsibility, ensuring that Europe's rich cultural collections contribute to knowledge advancement in ways that benefit society as a whole.
Readers interested in exploring the full document can visit the Europeana PRO announcement or the Open Future publication page. Additional resources on the common European data space for cultural heritage are available through official EU channels.
