OpenAI has banned a group of ChatGPT accounts originating in Russia that were used to support a covert influence campaign promoting a fictitious think tank and spreading narratives favorable to the Kremlin, the company announced Tuesday.
The operators routed their connections through VPNs to circumvent OpenAI's prohibition on model access from Russia. Working in Russian, they used ChatGPT to draft social media content distributed across Substack, Telegram, X, Facebook, and LinkedIn, while instructing the model to remove any traces that might reveal its AI origins. Most of the generated content was in English.
The campaign centered on an organization called the International Burke Institute, a self-described expert community with a listed street address in Israel. OpenAI said the site, registered in February 2025, falsely claimed to feature prominent academics including Francis Fukuyama and Noam Chomsky. Of 36 articles linked to experts on the site and published between September 2025 and May 2026, 34 were copied from elsewhere on the internet, sometimes with incorrect attribution.
A central feature of the operation was a "sovereignty index" that rated countries on a metric designed to cast Russia favorably and criticize Western nations. Reports targeting France, Germany, the United States and the European Union were described by OpenAI as veering into polemic.
The operation's direct reach was limited — most social media posts received low view counts, and the Telegram channels attracted between 10,000 and 20,000 followers each. However, OpenAI argued that the significance lay in the underlying apparatus: demonstrating how AI tools can help bad actors create an appearance of institutional credibility and lay groundwork for much larger operations.
