Microsoft has recently detected a network of social media accounts, suspected to be controlled by the Chinese, which are employing artificial intelligence (AI) technology to impersonate American voters and spread propaganda with the aim of influencing U.S. politics.
The detected network is similar to past operations linked to the Chinese Ministry of Public Security. What sets this operation apart is its use of generative AI since March 2023 to create content that mimics U.S. voters.
The AI-generated content is reportedly more eye-catching compared to previous campaigns by Chinese actors which relied on digital drawings, stock photo collages, and other manual graphic designs. Microsoft explains that these images are likely created by diffusion-powered image generators that utilize AI to not only generate compelling images but also improve them over time.
The influence network has been engaged in a campaign focusing on politically divisive topics such as gun violence and denigrating U.S. political figures and symbols. An example provided by Microsoft researchers was an AI-generated image of the Statue of Liberty brandishing an assault rifle with a caption saying, “Everything is being thrown away. THE GODDESS OF VIOLENCE.” Another example was a Black Lives Matter poster seemingly created by a “Chinese Communist Party-affiliated automated account” and later uploaded by an account impersonating a U.S. conservative voter.
While these social media posts may not seem particularly persuasive or well-disguised as Chinese propaganda, Microsoft emphasizes that AI generated these posts with minimal human intervention and it’s learning from its mistakes. This means it is building a more formidable propaganda operation as it progresses. Researchers have noted that the Chinese influence networks are improving their responses to human interaction.
A Microsoft spokesperson told Reuters that the China-linked accounts tried to appear American by listing their public location within the United States, posting American political slogans, and sharing hashtags related to domestic political issues. Microsoft’s threat report also highlighted some general observations about Chinese cyber activities, including the South China Sea, the U.S. defense industrial base, and U.S. critical infrastructure as the top three subjects of Chinese hacking activity in 2023.