OpenAI Bans Pro-Russia ChatGPT Accounts Tied to Israel-Based Think Tank Using Fabricated Researcher Credentials

Artificial intelligence leader OpenAI announced this week the successful disruption of a coordinated influence operation by banning a significant cluster of ChatGPT accounts. These accounts were allegedly used to disseminate pro-Kremlin narratives and promote the International Burke Institute (IBI), an entity described by OpenAI as the "core" of a Russian influence campaign, despite its self-proclaimed identity as an Israel-based think tank. The operation's methodology involved circumventing OpenAI's restrictions on users in Russia by employing Virtual Private Networks (VPNs), a common tactic in evading geo-blocking and censorship. Once access was gained, these users leveraged OpenAI's advanced language models to generate social media content designed to amplify pro-Russian viewpoints and bolster the IBI's profile.

The IBI, which maintains an active online presence, purports to be an "expert community." However, OpenAI's investigation revealed a pattern of deceptive practices. The think tank's website falsely listed prominent international figures, including renowned American linguist and philosopher Noam Chomsky and influential political scientist Francis Fukuyama, as contributors. Neither Chomsky nor Fukuyama immediately responded to inquiries from Fortune regarding their alleged association. Further investigation by Fortune uncovered that at least one author whose name was featured on the IBI's site was unaware of its use until contacted. The deception extended to listing deceased academics, such as Shlomo Avineri, an Israeli political scientist, and Jiang Ping, a Chinese legal scholar, as contributors. Notably, some of these individuals had passed away years before the IBI's reported founding date.

OpenAI's analysis also pointed to the IBI's website, registered in February 2025, containing a "sovereignty index" that demonstrably favored Russia. The platform hosted dozens of articles where authors were misattributed, adding another layer to the fabricated legitimacy. A review of the IBI's website archives by Fortune on the Wayback Machine indicated that while the site existed as of February of the preceding year, no content was published online until October 2025, suggesting a recent and potentially manufactured history. In response to initial reporting on OpenAI's findings by the French outlet Le Monde, the International Burke Institute issued a statement vehemently rejecting the characterization of its operations. The institute asserted that the mere use or citation of its publications or findings by third parties does not inherently prove cooperation, coordination, funding, or a shared underlying purpose. The IBI did not provide further comment when approached by Fortune.