An agent-based operationalization of DISARM for FIMI investigation on Social Media

Fernandez Toledano, J. (Co-author)
De Clerck, B. (Co-author)
Kevin Tseng (Co-author)
Phil Tinn (Speaker)
Yuliia Dukach (Co-author)

Activity: Talk or presentation › Oral scientific presentation

Description

Agent-based operationalization of the DISARM framework enables the systematic investigation of Foreign Information Manipulation and Interference (FIMI) on social media through the integration of Large Language Models, multi-agent systems, and statistical evidence validation. Specialized AI agents collaboratively identify candidate manipulative behaviors, map them to standardized DISARM tactics and techniques, and verify findings against ground-truth data in a transparent and reproducible manner. The approach enhances the scalability, explainability, and interoperability of FIMI analysis while supporting human-AI teaming, situational awareness, and intelligence-driven decision-making in military and security environments.

Period	12 May 2026 → 13 May 2026
Held at	International Conference on Military Communication and Information Systems (ICMCIS), organized by the Information Systems Technology (IST) Scientific and Technical Committee, IST-224-RSY, United Kingdom
Degree of Recognition	International

Keywords

Foreign Information Manipulation and Interference (FIMI)
DISARM Framework
Large Language Models (LLMs)
Multi-Agent Systems
OSINT
Information Operations (IO)