A research report from Clemson University revealed the presence of an AI-powered network of political propaganda accounts on X, posing as real people to support Republican candidates and causes. The network consisted of at least 686 identified accounts that made over 130,000 posts supporting specific Senate and primary races, as well as former President Donald Trump’s re-election campaign.
The accounts, using large language models (LLMs) similar to ChatGPT, engaged in coordinated interactions to push political narratives. Despite efforts to remove some accounts from X, the platform did not respond to inquiries from NBC News. The accounts strategically targeted popular posts to maximize visibility.
Researchers found evidence that the network utilized AI tools to generate and post content automatically. The accounts often included hashtags that were not commonly used by real individuals, signaling a possible AI origin. The use of bots and fake accounts on social media platforms has been a longstanding issue, but the adoption of LLM technology has enabled more convincing and interactive interactions.
The study suggested that the operation was likely American, given its focus on specific Republican campaigns. The tactics used by the network raised concerns about the potential for sophisticated disinformation campaigns in the future. While regulations on AI-driven propaganda are limited, the network’s activities raise questions about the need for increased oversight.
Overall, the report highlighted the growing threat of AI-driven disinformation campaigns and the challenges of combating such operations in a decentralized and rapidly evolving digital landscape. The use of advanced AI technology in shaping political discourse underscores the urgent need for robust strategies to address the spread of misinformation online.
Photo credit
www.nbcnews.com