Bridging Institutions, Wikidata, and Generative AI: From Data Silos to Shared Infrastructure
🎥 Session recording: [https://youtu.be/RWNgzZChMiA?list=PLfq0CVoHfK80&t=18855](https://youtu.be/RWNgzZChMiA?list=PLfq0CVoHfK80&t=18855)
Cultural heritage institutions, universities, research centers, governments and public organizations steward vast amounts of carefully curated data. Wikimedia projects, especially Wikidata and Wikibase, provide open, structured, multilingual infrastructures for connecting and sharing knowledge. At the same time, Generative AI systems are reshaping how information is queried, summarized, and reused. Yet these three domains, institutional data, structured open knowledge, and AI systems, often evolve in parallel rather than in coordination. This panel asks: how can institutional data be more sustainably connected to Wikidata and Wikibase, and how can structured, community-governed knowledge meaningfully inform AI systems?
Anchored in the theme of Freedom, Equity, and Reliability, the discussion will explore how lowering technical and organizational barriers can strengthen equitable participation by institutions of different sizes and regions; how structured, referenced data can contribute to more reliable AI outputs; and how open infrastructures can support institutional autonomy in an AI-shaped landscape. While the [AI-BRIDGES](https://ai-bridges.org) project will serve as one example of ongoing efforts to address these questions, offering practical insights from research and experimentation, the panel is designed as a broader conversation. Bringing together institutional partners, Open Knowledge experts and technologists, this panel will not only identify shared challenges and replicable pathways forward, but will engage the audience in expanding perspectives and ensuring inclusivity and equity around solutions.
Panelists include:
\* **Denny Vrandečić** (Wikidata founder; Abstract Wikipedia lead)
\* **Lydia Pintscher** (Wikidata)
\* **João Alexandre Peschanski** (WMBR)
\* **Pierre-Carl Langlais** (Pleias)
\* **Dee Harris** (Creative Commons)
