OpenAI disrupts Russian fake-think-tank influence campaign

OpenAI said it banned a cluster of ChatGPT accounts that "very likely originated in Russia" and were used to promote the International Burke Institute (IBI), a self-described "expert community" that claimed a street address in Israel. The associated website was registered in February 2025.

The operators prompted the model in Russian but generated mostly English-language social media comments, explicitly instructing ChatGPT to hide linguistic clues that they were Russian. Because OpenAI blocks access from Russia, they used VPNs. The resulting posts appeared on X, LinkedIn, Facebook, Substack and Telegram, some from accounts carrying the IBI name and logo, others from apparently inauthentic "everyday" accounts whose main activity was posting IBI articles.

The IBI site itself was not model-generated. In a sample of 36 articles attributed to IBI-listed experts and published between September 2025 and May 2026, OpenAI found 34 were copied from elsewhere — including a China-Pakistan Economic Corridor piece apparently lifted from Cambridge University Press and misattributed to a Nottingham professor, and a Migration Policy Institute article credited to an Australian food science professor. As of 17 July, the site listed luminaries including Francis Fukuyama and Noam Chomsky among its experts.

The site also published a "sovereignty index" rating countries in ways favourable to Russia. One operator generated German-language posts for a Telegram channel called "Lahme Ente" attacking Ukraine, the EU and the German government; another produced logos for a dozen Telegram channels targeting Germany, the US, France, Poland and Türkiye.

OpenAI assessed the campaign reached relatively small audiences. It did not name a sponsor or specify the number of accounts banned.