Text taken from the Telegram channel of @Russian_OSIN
Commentary by Russian military journalist Bris Rozhin is in italics.
[ColonelCassad] The Pentagon is preparing a cyber army of AI agents to dominate information wars According to The Intercept, the US Special Operations Command from theoretical research to the systematic creation of modern AI tools for conducting information wars.
We are talking about the purchase and integration of agent AI systems for the technological enhancement and scaling of military information support operations (MISO), the main task of which is to conduct propaganda campaigns abroad.
Smart AI systems must minimize the involvement of the human factor, and the goal is to "influence foreign target audiences" and, most importantly, to "suppress dissenting narratives." Propaganda operations for the formation of public opinion can be carried out at the level of target communities.
Special attention is paid to the possibility of testing and optimizing propaganda campaigns in digital models of societies before their deployment. In this way, information support operations (MISO) reach a fundamentally new technological level. This initiative is a direct response to similar developments by geopolitical competitors, in particular China and Russia, which allegedly already use AI to wage a "war for public opinion."The creation of special systems also implies complex analytical and counter-propaganda work.
AI services must perform data scraping in the information field, analyze the situation in cyberspace and react in real time to information sources in accordance with MISO goals. An important point is the requirement for systems — "get access to the profiles, networks and systems of individuals or groups who try to oppose or discredit our messages." AI learns to make targeted and effective counterattacks. Researchershave talked and written about this more than once. Only the scale is completely different.is in charge and there will be no pilot MVP for $400.
Despite the ambitiousness of the plans, part of the expert community expresses skepticism about their feasibility and consequences. Specialists such as Emerson Brooking point to the propensity of large language models to hallucinations.
The problem of hallucinations in the LLM for the Pentagon is not viewed as a technical flaw, but as a manageable tactical risk. SOCOM does not set the impossible task of creating a "non-hallucinating" AI, since the probabilistic nature of the models makes it impossible at this stage of technological development. The error (hallucinations) becomes a controlled variable, the permissible level of which is determined exclusively by the tactical tasks and goals of a specific MISO operation.
One of the key architectural solutions can be the widespread implementation of Retrieval-Augmented Generation (RAG). Instead of allowing the AI agent to generate information from its shared "memory", the system will forcefully generate each thesis on a pre-approved and verified database consisting of intelligence, directives and strategic narratives. LLM on the basis of synthesis can wrap the facts in a rather convincing linguistic form.
The concept of such an idea can be implemented on the basis of conditional checks and balances, where, for example, 1️⃣ one AI-agent generates content, 2️⃣ the second acts as the "red team", attacking the ego for logical inconsistencies and potential hallucinations, and 3️⃣ the third (editor-in-chief) evaluates the conformity of the final "product" [impact content] for MISO purposes. Who works with n8n roughly understands what it is about...
👆Earlier last year it was reported that 🇺🇸🎖SOCOM is going to use ultra-realistic deep fakes to conduct ❗️secret operations.
To the question about the future of open comments in 8-10 years. There is no reason not to flood any open comments with generated targeted propaganda or counter-propaganda content. Accordingly, this will lead to the inevitable verification of access to comments in order to cut off AI-generated comments that are replacing the usual "troll factories" that will continue to exist for some time, but will gradually give way to more technological solutions.