In the past, we searched for information in verified databases and books. These days, we tend to favour algorithms on social media platforms, search engines and artificial intelligence assistants developed by private companies that train their large language models on vast amounts of content (including digitized cultural heritage and copyright-protected works), often without the consent or remuneration of authors and memory institutions. Libraries and governments are actively trying to prevent human knowledge and its interpretations from coming under the control of a handful of commercial players. Where is the line between free use of work for research purposes and misuse of copyright-protected work to train commercial artificial intelligence? What is the stance of the German and Czech legislatures in relation to the EU AI Act? Is a library merely a place to borrow books, or is it a key component of an infrastructure of verification that ensures security and stability in society?
Frank Scholze (Deutsche Nationalbibliothek), Tomáš Foltýn (National Library of the Czech Republic), Achim Bonte (Staatsbibliothek zu Berlin) moderation: Saša Michailidis
