OpenAI Shuts Down Russian Accounts That Promoted a Fake “Institute”

OpenAI has reported that it blocked a cluster of Russian ChatGPT accounts that used a VPN to circumvent access restrictions. Using the AI tool, the operators generated social-media posts and comments that were then distributed on Substack, Telegram, X, Facebook and LinkedIn.

According to the company, the accounts promoted the so-called International Burke Institute (IBI) — an “expert community” formally based in Israel. What began as an investigation into AI-generated publications led to a far larger operation: a website with copied and falsely attributed academic works, a “sovereignty index” that cast Russia in a favorable light, and efforts to conceal the Russian origin of the authors.

Modest Reach, a Polished Storefront

The campaign's audience turned out to be small: the most visible assets were Telegram channels that had each gathered 10,000–20,000 subscribers. Most of the content was published in English, and the operators explicitly instructed ChatGPT to remove any linguistic traces that might give away the Russian-speaking authors.

The site ibi.institute was registered in February 2025 and presents itself as an association of leading experts in economics, politics, sociology and international relations. There are no signs that the articles were written via ChatGPT, but many of them appear to have been copied from scholarly publications — sometimes with false authorship — while others were written by a Slavic author and machine-translated into English.

The “Sovereignty Index”

OpenAI cites the operation's signature “sovereignty index” (also known as the Burke Index) as a defining feature. It is calculated across seven dimensions — political, economic, technological, informational, cultural, cognitive and military — summing scores in a range from 100 to 700.

The final country ranking is led by the United States (650.9), followed by China (649.1), Switzerland (610.7) and Russia (601.4). The report on Russia describes it as a state with a “high level of sovereignty,” record-setting military and economic self-sufficiency, but limited flexibility in its civilian economy; it separately notes “complete digital and military-nuclear sovereignty.”

The campaign's significance, OpenAI stresses, lies not in its reach but in the infrastructure it built. Individual promotional posts led to an outwardly convincing “institute” with fictitious experts, reprinted scholarly works and its own “risk index.” This shows how AI is becoming a supporting tool for fabricating authority and concealing the source of desired narratives — and how that same use of AI ultimately helps expose the entire operation.

Source: The Hacker News