Editorial illustration for Russian Influence Campaign Used ChatGPT to Push Pro-Kremlin Posts
Russian Operation Used ChatGPT for Kremlin Propaganda
Russian Influence Campaign Used ChatGPT to Push Pro-Kremlin Posts
OpenAI banned a cluster of ChatGPT accounts this year after tracing them to a covert Russian influence operation built around a fictitious think tank called the International Burke Institute. The people running it logged in from Russia through VPNs and told ChatGPT explicitly to scrub any language that might give away their location, according to OpenAI's account of the takedown.
The institute, which claimed to be based in Israel, put out a "sovereignty index" ranking countries in a way that flattered Russia and made Western democracies look worse by comparison. OpenAI found that 34 of 36 articles on the site attributed to named experts between September 2025 and May 2026 were lifted from other publications, in some cases with invented author credentials. One piece taken from Cambridge University Press ended up credited to a University of Nottingham professor who had nothing to do with it. Another, sourced from the Migration Policy Institute, was pinned on an Australian food chemistry academic.
The operation didn't stop at one website. It fed into a wider network of social accounts and Telegram channels stretching across several countries, which is where the scale of the effort starts to show.
OpenAI has exposed a covert Russian influence campaign that used ChatGPT to generate social media posts and disrupted it by banning a cluster of ChatGPT accounts.
Why this matters
OpenAI's takedown shows the guardrails work at least once, but the pattern is the part worth watching. A handful of operators in Russia used VPNs and plain instructions to strip out linguistic tells, then farmed the output across five platforms in two languages, from a "think tank" front to a German Telegram channel called "Lahme Ente." None of that required custom models or jailbreaks, just a free account and patience. For developers and founders building on these APIs, the lesson isn't that ChatGPT is uniquely exploitable, it's that any general-purpose model becomes cheap infrastructure for propaganda the moment someone scripts around it.
Detection here relied on OpenAI's own visibility into account behavior, not on the content itself looking fake, which is a narrower defense than it sounds. Researchers should treat "low reach" campaigns like this one as dry runs, not failures. The interesting question now is whether platforms downstream, X, LinkedIn, Substack, are watching for these clusters too, or waiting for OpenAI to keep doing the flagging for them.
Related Reading
Browse all OpenAI news →
LLMs & Generative AI ChatGPT's 'Nerdy' tweak rewards goblin metaphors in answers, study finds
LLMs & Generative AI Google tests visual 'magazine-style' UI for Gemini 3 Pro users
LLMs & Generative AI AI Engineers Face Rising Costs, Need New Strategies for Efficiency
Related Topic OpenAI's GPT Transcribe Cuts Error Rate to 3.31%, Improving on GPT-4o
Open Source
Related Topic OpenAI Calls Apple's Lawsuit 'Rotten to Its Core' in Rebuttal
Policy & Regulation
From Archives Anthropic's Claude update ends need to repeatedly rebrief the AI
From Archives Meta Tests MetaRoCE on 64-Node AMD Cluster, Outperforms RoCEv2
From Archives OpenAI's Jalapeño Chip Boasts High Efficiency, Low Latency in Benchmarks
Research & Benchmarks
From Archives Alabama AG Subpoenas OpenAI Over Safety Practices Following Hack
Open Source
Common Questions Answered
What was the International Burke Institute and how did OpenAI discover it was part of a Russian influence operation?
The International Burke Institute was a fictitious think tank used as a front for a covert Russian influence campaign that claimed to be based in Israel. OpenAI traced ChatGPT accounts back to operators logging in from Russia through VPNs and discovered they were explicitly instructing ChatGPT to remove language that would reveal their geographic location, leading to the ban of the account cluster.
How did the Russian operators use ChatGPT to conduct their influence campaign across multiple platforms?
The operators used free ChatGPT accounts to generate social media posts promoting pro-Kremlin narratives, which they then distributed across five different platforms in two languages, including channels like the German Telegram account called 'Lahme Ente.' This approach required no custom models or jailbreaks, just basic instructions and patience to farm the AI-generated output across their network.
What does OpenAI's takedown of this Russian influence operation reveal about AI safety and future threats?
While OpenAI's action demonstrates that safety guardrails can work to detect and disrupt malicious campaigns, the incident reveals a concerning pattern: operators only needed VPNs, plain instructions, and free accounts to conduct a sophisticated multi-platform influence operation. This suggests that developers and founders building on AI APIs should be aware that relatively simple tactics can still be leveraged for large-scale coordinated inauthentic behavior without requiring advanced jailbreaks or custom models.
What was the 'sovereignty index' that the International Burke Institute published?
The International Burke Institute put out a 'sovereignty index' that ranked countries in a way designed to favor pro-Kremlin perspectives and narratives. This index served as part of the broader disinformation strategy to push geopolitical messaging aligned with Russian interests across Western audiences.
Further Reading
- OpenAI bans Russian ChatGPT accounts used in misinformation campaign - CNBC
- OpenAI banned Russian ChatGPT accounts used to spread pro-Kremlin misinformation - Quartz
- Russian disinformation on ChatGPT: Fake Telegram channels, fake videos and a real-fake Israeli think tank - Le Monde
- Disrupting a new covert influence campaign from Russia - AI Weekly
- OpenAI Bans Russian ChatGPT Cluster Behind 'International Burke ... - AI Weekly
