Blocking a New Covert Influence Campaign Originating from Russia
Key point
OpenAI has blocked covert influence campaigns exploiting ChatGPT that originated from Russia.
Details
OpenAI blocked clusters of ChatGPT accounts originating from Russia. These accounts accessed the platform via VPNs to promote an Israel-based "expert community" called the International Burke Institute (IBI).
Operators input prompts in Russian to generate English social media comments, instructing ChatGPT to hide traces of Russian. The generated content was posted on various platforms, including X, LinkedIn, Facebook, Substack, and Telegram.
The IBI website was registered in February 2025 and contained manipulated information, such as unauthorized copying of actual academic papers or citing false sources. Notably, unnatural expressions that appeared to be machine-translated from West Slavic languages (e.g., 'Svetofor coalition') were discovered.
The core of this campaign was the creation of a 'sovereignty index' that portrayed Russia positively and disparaged Western nations. OpenAI assessed this as the most sophisticated campaign among those it has blocked involving influence operations linked to Russia since the war in Ukraine.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.