Russia has begun using artificial intelligence-based bots for spreading propaganda on social media, especially on Telegram, according to a joint investigation by OpenMinds and the Digital Forensic Research Lab (DFRLab).
Researchers have uncovered over 3,600 bots that posted more than 316,000 AI-generated comments in Telegram channels linked to Ukraine's temporarily occupied territories. Another three million messages were spread in broader Ukrainian and Russian Telegram groups. These bots used human-like language, adapting replies to the context of each conversation to promote pro-Kremlin narratives and undermine Ukraine.
Unlike traditional bots that spam identical messages, these accounts simulate real users. They reply directly to other users, shift tone and content, and tailor messages to appear authentic. On average, a bot posts 84 comments per day, with some exceeding 1,000 daily.
The goal is not just to spread fake news, but to create the illusion of widespread public support for the occupation regime, filling comment sections with praise for Russia and attacks on Ukraine. In an environment of information isolation, this becomes a potent tool of mass manipulation.
AI-generated bots often give themselves away through:
- absurd usernames,
- unnatural or AI-generated profile pictures,
- overly formal or awkward phrasing,
- and highly diverse language: one in three comments is uniquely generated by AI.
Even when bot accounts are deleted, their influence lingers. Locals repeatedly exposed to these comments may perceive Kremlin propaganda as the majority opinion, especially in regions where Ukrainian news is inaccessible.
Read also
-
From making interceptors to guarding sky: Ukrainian drone maker gets its own private air defense unit and will pay for it from weapon sales
-
Russia burned 628,000 new school textbooks in Kharkiv at the peak of back-to-school season
-
Ukraine wants to aim Starlink at Russian targets—and build a version Elon Musk can’t switch off
