The Russian disinformation network Matryoshka launched a campaign using AI voice cloning tools to create fake videos of university professors making pro-Russian statements, spreading these deepfakes across social media platforms including X and Bluesky.
The Russian disinformation network Matryoshka, which began operations no later than September 2023, launched a new campaign using artificial intelligence voice cloning tools to create fake videos of university professors. The campaign targeted academics from prestigious institutions including Cambridge, Harvard, Princeton, University of Bristol, University of Cumbria, and Sciences Po. The videos begin with genuine footage of professors introducing themselves, but then transition to AI-generated audio segments where the cloned voices make pro-Russian statements, urge Ukraine to surrender territories, call for lifting sanctions against Russia, and even portray Ukrainian President Zelensky as a vampire. The University of Bristol verified the AI manipulation in a video featuring Professor Ronald Hutton. The fake videos are distributed by hundreds of bot accounts with minimal followers across social media platforms, reaching hundreds of shares on X and expanding to Bluesky with multilingual content in English, Dutch, Spanish, Indonesian, Thai, and Portuguese. French government agency Viginum has linked Matryoshka to the Kremlin-affiliated Doppelganger network run by Russia's GRU military intelligence. The campaign spreads anti-Ukrainian propaganda and disinformation aimed at manipulating public opinion about the war in Ukraine.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to conduct large-scale disinformation campaigns, malicious surveillance, or targeted and sophisticated automated censorship and propaganda, with the aim of manipulating political processes, public opinion, and behavior.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed