Building Bridges for the Future of Knowledge: How AI-BRIDGES Connects Data, AI, and Open Infrastructures

Generative AI is transforming the way we find, use, and understand knowledge. At the same time, many high-quality datasets from the fields of culture, research and education remain difficult to access or are rarely reused. This is precisely where AI-BRIDGES comes in: the research project connects institutional data, Open Knowledge infrastructures, and AI systems, with the goal of developing sustainable and responsible solutions for a better knowledge ecosystem.

AI-BRIDGES is being implemented at the Digital Humanities Research Hub, School of Advanced Study, University of London and is funded by the European Commission as part of a Marie Skłodowska-Curie Postdoctoral Fellowship. This is a two-year project, from September 2025 to August 2027, and is designed from the outset so that its outcomes are sustainable long after the project’s duration. To ensure its success, the project consists of a network of experienced partners.

AI-BRIDGES is being developed in close collaboration with an international network of 5 key partners: Wikimedia Deutschland, as a partner for Wikidata, Wikibase, and open data infrastructures; Wikimedia UK with a focus on institutional collaborations and public engagement; Wikimedia Brazil with strong expertise in Wikidata & Wikibase, community building, and working in less-resourced countries, and an inclusive, non-English-centric approach; Wikimedia Switzerland with a focus on coordinated infrastructure development; and Pleias, the DeepTech startup and AI lab, with expertise in cultural heritage, open source language models and AI systems for the public sector.

With its many years of experience with open data and technologies such as Wikidata and Wikibase, Wikimedia Deutschland provides the central infrastructure through which massive, verified, crowdsourced data can find its way onto the internet. In addition, the German association already has experience in building bridges through the Wikidata Embedding Project, where data from the Wikidata knowledge graph was embedded as a vector database, so that it could be more easily utilized by AI systems. The primary goal of the project is to give the open-source community the opportunity to develop open AI applications more easily.

In addition to these core partners, AI-BRIDGES collaborates with a growing network of individuals, communities, organizations, and institutions that contribute their time, expertise, and perspectives to various areas of the project.

To this end, the project develops workflows, tools, and collaborative processes that make it easier for institutions to structure, share, and reuse their data. A central component of this is Wikidata and the Wikibase ecosystem, an open infrastructure that allows data to be modeled and be publicly accessible for both humans and machines.

At the same time, Wikimedia Deutschland is researching how generative AI can meaningfully access structured, open data. Rather than viewing generative AI as an isolated black box, the goal is to connect it with reliable data sources using transparent and traceable methods in the public interest.

AI-BRIDGES is designed as a research project, with the goal of producing results that will have an impact beyond the project’s duration. These include, among other things:

• Assisting institutions share their data openly in an easier, responsible, and more sustainable way.
• Approaches to how Open Knowledge infrastructures and AI systems can collaborate more effectively.
• Participatory models that also involve communities and the public in data work. The key term here is digital literacy.
• Concrete tools, workflows and datasets that can be reused in various contexts.
• The prospective development of an open, structured-data-aware language model in collaboration with Pleias.

The goal is a networked ecosystem in which data does not remain isolated, but can be meaningfully used, understood, and further developed.https://www.youtube.com/watch?v=pFvjU_SRhK0

Amid the current generative AI boom, people often overlook where knowledge actually comes from. Many systems draw on existing content without revealing its source or integrating it sustainably.

AI-BRIDGES aims to change this: It creates connections between the places where knowledge is generated and the technologies that disseminate it. In this way, the project helps to strengthen transparency, traceability and accountability in the handling of data. In other words, if generative AI is the future of access to knowledge, then the foundation upon which this knowledge rests is being decided right now.

Institutions, researchers, developers, communities, and funders are invited to get involved – whether through data donations, expertise, or collaborative projects.

Especially during this initial development phase, the goal is to bring together as much data as possible within open and free infrastructures. Wikidata and the Wikibase ecosystem enable data to be captured in a structured way and made publicly accessible, an important foundation for its future use, including in the context of generative AI.