ChatGPT has blocked accounts from Russia linked to disinformation
OpenAI has suspended a group of ChatGPT accounts linked to a covert political disinformation campaign. According to the company, users from Russia and, likely, Israel were involved in the operation, using a chatbot to prepare content for social media.
The campaign participants promoted the fictitious International Burke Institute (IBI), which positioned itself as an expert community in Israel and published content about the politics of various countries.
“This is the first time we have disrupted a Russia-linked influence operation that has gone this far,” OpenAI stated.
The International Burke Institute website was registered in February 2025. On the site, the IBI was presented as a research center and expert community.
Among the purported experts listed were American political scientists Francis Fukuyama and Noam Chomsky. However, according to OpenAI, the information about their affiliation with the institute was false.
A review of the materials revealed that 34 of the 36 publications on the IBI website were copied from other sources. Moreover, in most cases, the authors listed were incorrect.
According to OpenAI, this practice was intended to create the appearance of a legitimate research center and lend greater credibility to its publications.
According to OpenAI, the participants in the operation connected to ChatGPT via a VPN and submitted queries in Russian.
At the same time, they created most of the content in English. Users also asked the model to remove linguistic features that might indicate the Russian origin of the texts.
Thus, ChatGPT was used as a tool to prepare and adapt content for a covert information campaign aimed at creating the illusion of an independent think tank.
This was reported by OpenAI.
Previously, OpenAI introduced ChatGPT for Teens—a separate chatbot format for users under the age of 18 with enhanced built-in safety features. It combines educational tools with age-based safeguards designed to reduce the risk of teenagers encountering content that is harmful or inappropriate for their development, according to CNBC.
