How Generative AI Is Being Used and Confused by Inauthentic Influence Campaigns
Alethea identified a network of at least 400 X accounts exhibiting inauthentic behavior and coordinated narrative promotion. The network used large language models (LLMs) to generate subtly varied copypasta-style content—a technique Alethea terms PromptPasta—to amplify messages aligned with key figures and policy positions associated with the second Trump Administration. The outputs of PromptPasta are nuance
.png)