This post is also available in:
עברית (Hebrew)
Modern military operations are influenced by more than events on the physical battlefield. Social media posts, messaging platforms, videos, news reports and foreign-language sources can shape public perception and provide clues about emerging threats. The problem for intelligence teams is scale: analysts may need to monitor millions of pieces of content across hundreds of sources and languages while identifying which developments actually matter.
The U.S. Marine Corps is adopting an AI platform designed to accelerate that process. Argus for Cognitive Advantage, developed by Accrete, analyzes open-source information to identify narratives, track how they evolve and turn large volumes of online content into information analysts can investigate.
According to Interesting Engineering, the system collects information from the open web, social networks, broadcast media, messaging services and foreign-language outlets. Its AI agents then search those sources for signals relevant to specific missions rather than requiring analysts to manually sift through every post or report.
The system is built on the company’s Knowledge Engine, which is designed to encode the expertise of experienced personnel into specialized AI agents. Those agents can identify relevant information, connect related developments and generate reports intended to shorten the time between discovering an event and understanding its significance.
Video receives particular attention. The system can automatically transcribe and translate speech, identify objects and scenes, and extract clues about where footage may have been recorded. Long videos can be reduced to relevant segments, while information contained inside them becomes searchable rather than requiring analysts to watch hours of footage manually.
The platform also includes narrative and stance detection, allowing users to examine how particular conversations develop and how different audiences respond. An Audience Simulation Agent can divide audiences into groups and model how each might react to a proposed message before it is released.
For military information operations, that could help analysts understand how adversarial narratives are spreading, identify which communities are engaging with them and evaluate potential responses. Predictive simulations are intended to support planning, while automated assessments can help measure the results of an information campaign afterward.
The system supports real-time transcription and translation across 243 languages and can analyze sources including Telegram, TikTok, YouTube, Reddit, X, Weibo and VKontakte, among others. Analysts can also interact with collected information through Argus Chat, asking questions in natural language.
An important feature for intelligence work is traceability. The company says AI-generated conclusions can be linked back to their original sources, allowing analysts to examine the evidence rather than relying solely on an automated assessment.
The Marine Corps’ adoption reflects the growing role of AI in information warfare. The advantage is not simply collecting more data, but identifying relevant signals quickly enough for human analysts to understand and act on them before the information environment changes again.
