Defending American Interests Abroad: Early Detection of Foreign Malign Information Operations

Mouton, Lucas, and Ee's 24-page report represents RAND's first systematic investigation of using large language models for early detection of foreign malign information operations in overseas media environments. The methodology applies LLMs to identify classical propaganda techniques — appeal to fear, scapegoating, false dilemmas, bandwagon effects — in media content targeting American interests abroad, providing a first-step detection signal before more resource-intensive human analysis.

The approach is significant for what it does and does not claim. The authors are explicit that LLM-based detection is a triage tool — it identifies content for further investigation, not a definitive attribution system. This epistemic modesty is appropriate and important: automated detection of propaganda techniques is a tractable problem; automated attribution of information operations is not. For practitioners building scalable FIMI detection infrastructure, this report provides the most recent RAND methodology on AI-assisted detection and a frank assessment of where the technology is reliable and where it is not.