Austrian Federal Ministry of Defence (BMLV): IBM watsonx-Powered 'ChatZentDoc' OSINT Tool for Daily Intelligence Analysis

Title

Austrian Federal Ministry of Defence (BMLV): IBM watsonx-Powered 'ChatZentDoc' OSINT Tool for Daily Intelligence Analysis

Content

Austria's Federal Ministry for National Defence deployed "ChatZentDoc," an OSINT (Open-Source Intelligence) tool built on IBM watsonx.ai that helps defense analysts query and classify a daily-updated stream of open-source information using natural-language questions and exportable, source-attributed answers. "The quality of the first presentation positively surprised us all. No project has worked so quickly so successfully," said Colonel Klaus Mak, who leads ZentDok, the ministry's central documentation service. The deployment was driven by an information-overload problem. Defense analysts can observe data floods with human intelligence but cannot assess them at scale — yet their evaluation and classification is the prerequisite for remaining capable of acting under risk scenarios. AI solutions for the public sector posed a particular compliance challenge, given GDPR, the IT Security Act, and the EU AI Act. The technical foundation was built on a long-standing relationship. ZentDok has used IBM Watson Explorer for natural-language processing since 2016 and built on that experience with IBM watsonx.ai AI studio for the new ChatZentDoc. The interface lets users query the database in natural language, isolate answers with filter options, and export responses along with source documents. The system is fed by information updated daily to ensure timeliness, and is positioned as an OSINT pattern that other public-administration applications could reuse. Benedikt Klotz of IBM framed the broader applicability: "Open-source intelligence is universally usable in areas that need to search and analyze quantities of unstructured data. AI solutions for the public sector always posed a particular challenge in terms of security and regulatory provisions such as GDPR, the IT Security Act and the EU AI Act. The experiences gained in this pilot also benefit other business areas within the economy that deal with sensitive data." The pilot was developed through the IBM Client Engineering method, which gives customers hands-on access to products under expert guidance. Looking ahead, the next step is a test environment in which users of other departments can exercise the tool, paired with training in generative AI and prompt engineering. "We are entering a completely new world here," Mak said. "User prompting is the essential thing, and the user's capability will become more and more important."