Kathrine Frich
The Pentagon envisions systems that can track online discourse, identify dissenting voices, and counter them with tailored messaging.
The Pentagon envisions systems that can track online discourse, identify dissenting voices, and counter them with tailored messaging.
Pentagon Eyes AI-Powered Propaganda to Sway Global Opinion
A newly revealed Pentagon document suggests that the United States is exploring the use of AI-generated propaganda to influence foreign populations.
The plan, disclosed by The Intercept, outlines the military’s intent to use AI to “control narratives” and “suppress dissenting arguments” online in real time.
SOCOM Plans ‘Autonomous’ Influence Campaigns
The document, created by U.S. Special Operations Command (SOCOM), calls for AI agents and large language model (LLM) systems that can scale influence operations with little human oversight.
SOCOM admits that the pace of digital information requires faster, automated responses.
“A program built to support our objectives can allow us to control narratives and influence audiences in real time,” the document states.
Software Could Target Individuals and Suppress Dissent
The Pentagon envisions systems that can track online discourse, identify dissenting voices, and counter them with tailored messaging.
SOCOM seeks “automated systems that extract information from the online environment, analyze the situation, and respond with messages consistent with its objectives.”
Propaganda Could Be Tailored to Each Target
The Pentagon hopes to build systems that can access and analyze the profiles and networks of individuals or groups who criticize U.S. narratives.
The aim is to create targeted messages “capable of influencing that specific individual or group,” the document reveals.
Simulating and Deploying AI-Driven Messaging
SOCOM also wants tools that can simulate how a propaganda message might be received by the public before it is released.
The Pentagon anticipates using generative AI systems not only to design content but also to test its psychological impact on different audiences.
Pentagon Also Exploring Deepfakes for Offensive Use
Among the tools listed in the Pentagon’s so-called “wish list” is deepfake technology for potential offensive capabilities.
This includes generating fake images or video that could support broader influence campaigns, a tactic often criticized when used by other regimes.
Critics Warn of ‘Repressive Power’ and Global Impact
A 2023 Freedom House report funded by the U.S. State Department warned of the “repressive power of artificial intelligence”.
The report predicted a rise in AI-driven disinformation and cited concerns over how authoritarian governments might exploit such technologies.
Officials Say U.S. Must Compete With China and Russia
William Marcellino of RAND Corporation said such efforts are necessary to compete with rivals.
“Regimes like those in China and Russia are engaged in AI-powered, large-scale malign influence efforts,” he said. “Countering these campaigns likely requires AI responses at scale.”
Rules Ban Domestic Propaganda, but Loopholes Remain
Though U.S. law prohibits propaganda aimed at Americans, the anonymous nature of online platforms makes enforcement difficult.
A SOCOM spokesperson insisted that the programs are aligned with “U.S. law and policy” and said that all efforts fall under the Department of Defense’s “Responsible AI” framework.
