OpenAI Removes Russia-Linked Accounts Promoting IBI Network
OpenAI has banned a cluster of ChatGPT accounts that it says very likely originated in Russia and were used to promote the International Burke Institute (IBI), a self-described expert community claiming to be based in Israel. The operators used the service to generate social-media content for X, LinkedIn, Facebook, Substack and Telegram, while the associated operation built a website, purported expert network and a “sovereignty” index that presented Russia favourably.
OpenAI assessed the campaign’s immediate impact as limited: typical social posts received low view counts and the official IBI accounts had few subscribers. Its Telegram channels had greater reach, generally attracting 10,000 to 20,000 followers each. On the Brookings Breakout Scale, OpenAI placed the activity at the lower end of Category Three, covering multi-platform activity with some signs of breakout to authentic audiences.
ChatGPT used for promotion rather than the core website
The account operators prompted in Russian and generated mostly English comments, explicitly instructing ChatGPT to avoid linguistic signs that they were Russian. Because OpenAI does not permit model access from Russia, the operators used VPNs. The generated material promoted IBI articles, including posts from accounts using the institute’s name and logo as well as accounts that appeared to be ordinary users.
The group also produced replies to genuine Substack users, usually asking them to follow the IBI channel. One operator created German-language posts for the Telegram channel Lahme Ente, which routinely criticised Ukraine, the European Union and the German government while advocating closer relations with Russia. Another generated logos for a dozen Telegram channels focused on Germany, the United States, France, Poland and Türkiye, and repeatedly requested Russian-language summaries of their activity.
Copied research and a claimed sovereignty index
IBI’s website was registered in February 2025 and gave an Israeli street address. OpenAI said material hosted on the site was not generated with its models. In a sample of 36 articles linked to experts and published between September 2025 and May 2026, 34 had been copied from elsewhere online. Some were old works or were attributed to the wrong authors.
OpenAI cited an article on the China-Pakistan Economic Corridor that appeared to reproduce a Cambridge University Press publication while crediting a University of Nottingham professor. It also identified migration-governance material apparently copied from the Migration Policy Institute and attributed to an Australian professor of food science. The site claimed prominent experts, including Francis Fukuyama and Noam Chomsky.
Alongside the republished work, IBI issued country reports framed around its sovereignty index. The reports addressed countries including France, Germany and the United States, as well as the European Union, and used rhetoric critical of Western governments. OpenAI said the campaign’s more notable feature was the effort to manufacture an apparently credible institution with named experts, research material and a purported proprietary index.
Infrastructure is the central risk
OpenAI said it could not determine the relationship between several real individuals in Israel who may have represented IBI in submissions or at conferences, the institute and the Russia-based operators. Its disruption focused on the ChatGPT accounts used to generate promotional content, rather than claiming the website itself was AI-generated.
For businesses, researchers and communications teams, the practical implication is to verify authorship, institutional affiliations and the origins of claimed proprietary indicators before treating online research or expert-branded content as reliable evidence.