A feasibility study for a European Books Data Commons
This week the National Library of the Netherlands (KB) and the Europeana Foundation published a feasibility study for the European Books Data Commons (EBDC). The EBDC is an initiative to make the full text of millions of digitized books held by European libraries available for research, innovation and the development of trustworthy AI applications. It builds on existing digital library collections and would turn public domain books into a high-quality, multilingual, AI-ready dataset that can be reused under clear and transparent conditions. It is designed to operate within the common European data space for cultural heritage, providing the infrastructure through which cultural heritage institutions can make large-scale datasets available for reuse. In doing so, the EBDC aims to contribute to a European Public AI infrastructure built on public values.
From outline to feasibility study
The initiative traces back to our Outline for a European Books Data Commons, published in November 2025. That concept paper proposed the EBDC as public digital infrastructure that would give researchers and AI developers a single point of access to the digitized public domain collections of European libraries, which today remain largely locked away and hard to use.
The feasibility study takes that concept and tests it against reality. Commissioned by KB and Europeana and carried out by Open Future, it assesses the technical, legal, organizational and financial requirements for establishing the EBDC. Its conclusion is clear: it is feasible to build this European resource under the stewardship of libraries.
What comes next
The next step is a small-scale pilot. A group of libraries will be invited to contribute public domain collections, technical expertise and governance knowledge, and to test the proposed approach in practice. The results will give other libraries the information they need to decide whether to join a later phase.
For us this is an encouraging milestone. We have long argued that European cultural heritage institutions should not simply be sources of training data for others, but should have a hand in shaping how knowledge and culture are made available in the age of AI. As KB’s General Director Wilma van Wezenbeek puts it, the EBDC “offers a unique opportunity to build an AI-ready corpus on our own terms.” A commons of digitized books, governed by libraries and organized in the public interest, is a concrete way to put that principle into practice, and a building block for a Public AI ecosystem in Europe.
We want to thank KB and Europeana for backing the idea and for their commitment to carrying it forward. We look forward to shaping the pilot phase, and to seeing the EBDC take shape.
