OpenAI Bans Russian ChatGPT accounts participating in a covert online influence operation aimed at manufacturing authority and promoting narratives favorable to Moscow.
The campaign leveraged generative AI tools alongside an elaborate network of fake digital infrastructure, including a front think tank, plagiarized academic work, and a biased geopolitical index.
Inside the “International Burke Institute” Operation
The operators created a fake front organization named the International Burke Institute (IBI), which falsely claimed to be an Israel-based expert research community.
- Plagiarized Expertise: A review of 36 articles published on the IBI website between late 2025 and May 2026 revealed that 34 were directly plagiarized or misattributed. In one instance, a paper on migration policy was falsely attributed to an Australian food science professor.
- Biased “Sovereignty Index”: The group created a custom ranking metric designed to portray Russia positively while systematically denigrating Western nations, the EU, Germany, France, and the United States.
- AI-Assisted Content Creation: The threat actors submitted prompts in Russian to instruct ChatGPT to generate social media comments, articles, and Telegram channel logos in English and German.
OpenAI stated,
“We recently banned a cluster of ChatGPT accounts originating in Russia that were being used to promote the International Burke Institute (IBI), a self-described “expert community” based in Israel. What began as an investigation into AI-generated social media posts led us to a much broader influence operation, built around a website containing copied and misattributed academic work, a “sovereignty” index that cast Russia in a favourable light, and efforts to disguise the operators’ Russian origins.”
Evasion Tactics and Technical Execution
To conceal their operational footprint and bypass safety guardrails:
1. Linguistic Laundering: Prompts explicitly instructed ChatGPT to rewrite content to eliminate Slavic phrasing and speech patterns that could betray the operators’ Russian origins.
2. Bypassing Geo-Restrictions: Because OpenAI blocks access to its services from Russia, the operators routed traffic through Virtual Private Networks (VPNs) to hide their true IP locations.
3. Multi-Platform Deployment: Generated text was pushed across Substack, Telegram, X (formerly Twitter), Facebook, and LinkedIn.
Impact and Industry Takeaways
OpenAI noted that while the operation achieved limited engagement—drawing minimal views and low follower counts—the sophistication lay in its architectural design. Instead of relying solely on automated bot spam, the actors built an apparently legitimate institution to obscure source origins and manufacture authority over time.
The bans mark the latest enforcement action by OpenAI against state-aligned actors, following previous disruptions of Russian-linked groups (such as Rybar) abusing LLM infrastructure for cyber operations and propaganda creation.
