Russia has launched a large-scale digital influence operation codenamed “Matryoshka,” using artificial intelligence to turn materials from the Jeffrey Epstein case into a powerful information weapon, reports EdgeTheory.
According to a detailed report by analytics firm EdgeTheory, this campaign marks a new milestone in disinformation tactics. The network gets its name from its unique multi-layered structure: bots on the social media platform X (formerly Twitter) do not simply post fakes but instead direct users through a chain of intermediary sites that mimic legitimate Western media outlets. AI enables the system to generate an endless stream of unique articles and deepfake videos in which U.S. politicians and opinion leaders are falsely accused of sexual crimes. The main strategy is to flood the information space with fake court documents so well-produced that ordinary users can hardly distinguish them from real ones.
Technical analysis of “Matryoshka” shows the operation relies on advanced large language models (LLMs), allowing bots to engage in coherent dialogues in multiple languages, including English, French, and German. Unlike primitive bot farms of the past, these accounts can mimic individual communication styles, participate in debates, and respond to criticism, creating a strong illusion of reality. EdgeTheory emphasizes that the campaign exploits a “plausibility effect”: fake lists of names from the Epstein case are mixed with real facts, causing audiences to believe the entire false narrative. The network also uses automated voice-cloning technology to create fake podcasts and news briefs voiced by well-known journalists, distributed via private communities and messaging apps to bypass standard social media filters.
A particularly dangerous aspect is real-time integration of deepfakes. Researchers documented instances where “Matryoshka” responded to current events within hours, producing generated video commentary from “eyewitnesses” or “victims” shortly after real news reports. This indicates powerful computing infrastructure and direct state funding. The operation aims not only to discredit specific individuals but to generate total cognitive chaos in Western societies ahead of major political events in 2026.
EdgeTheory analysts warn that such automated disinformation renders traditional fact-checking methods almost useless, as the speed of fake content generation far exceeds the pace of its debunking.
