---
title: "Open Data/Freie Daten in KI Chatbots nutzen"
license: CC-BY-4.0
license_url: http://creativecommons.org/licenses/by/4.0
license_note: "The recording page https://media.ccc.de/v/clt25-269-open-datafreie-daten-in-ki-chatbots-nutzen states: Licensed to the public under http://creativecommons.org/licenses/by/4.0. Confirmed 2026-10-18. The same sentence is in the event description returned by the public media.ccc.de API. No speech-subtitle track was listed. This note follows that description. It is an original summary, not a transcript."
source: https://media.ccc.de/v/clt25-269-open-datafreie-daten-in-ki-chatbots-nutzen
source_license: "CC BY 4.0"
attribution: "Summary by iinventedlight (AI-assisted). Source: Michael Christen, Open Data/Freie Daten in KI Chatbots nutzen, Chemnitzer Linux-Tage 2025 (CC BY 4.0)."
date: 2026-10-18
status: published
category: talks
tags: open data, retrieval, chatbots, language models
checked_on: 2026-10-18
sources:
  - id: S1
    title: "Open Data/Freie Daten in KI Chatbots nutzen"
    authors: "Michael Christen"
    url: https://media.ccc.de/v/clt25-269-open-datafreie-daten-in-ki-chatbots-nutzen
    accessed: 2026-10-18
    license: CC BY 4.0
    attribution: "Michael Christen, Open Data/Freie Daten in KI Chatbots nutzen, Chemnitzer Linux-Tage 2025, CC BY 4.0, http://creativecommons.org/licenses/by/4.0, https://media.ccc.de/v/clt25-269-open-datafreie-daten-in-ki-chatbots-nutzen"
---
# Open Data/Freie Daten in KI Chatbots nutzen

License: [CC BY 4.0](http://creativecommons.org/licenses/by/4.0). Source: https://media.ccc.de/v/clt25-269-open-datafreie-daten-in-ki-chatbots-nutzen

## Summary
Michael Christen, speaking at the Chemnitzer Linux-Tage on 22 March 2025, shows how a chatbot can answer from open data instead of from its training cutoff alone. The description says generative models and large language models are trained on huge collections, and that current data and niche topics are still missing from that training. The method he names is retrieval-augmented generation. He says a live demo will show open data, for example laws and administrative data, wiki dumps, ZIM files, and web crawls, and also a holder's own data, for example company data, flowing into the chatbot's replies. The page names the method and the sources. It does not define the retrieval step. The talk continues his 2024 Chemnitzer Linux-Tage talk on how ChatGPT works. The stated aim is that the open-source community should understand how complex AI systems work, so those systems can serve a culture of free data and decentralized systems. This note follows the description on the recording page. The page lists no speech subtitles, so it does not report the demo's steps or results.

## Why it is here
Askers who want to try an assistant on a library's own material need a way for the answer to come from that material. This talk's description is a practical account of retrieval-augmented generation over open data, including laws, wiki dumps, and web crawls.