OpenAI disables Russian-linked ChatGPT accounts promoting dubious Israeli think tank

OpenAI announced it had banned a cluster of ChatGPT accounts that evaded its restriction on Russian users by employing VPNs, using the model to produce posts that criticized Western nations and highlighted the International Burke Institute (IBI). The IBI, which presents itself as an Israeli expert community, is described by OpenAI as the centerpiece of a Russian influence operation. Its website claims contributions from figures like Noam Chomsky and Francis Fukuyama, yet neither has responded, and some listed experts died before the institute’s 2025 registration.

An audit found that 34 of 36 articles were authentic scholarly pieces incorrectly credited to unrelated authors, including a migration-policy paper misattributed to Kate Howell. Social-media output from the operation garnered modest engagement on platforms such as Facebook and X, but its Telegram channels drew between 10,000 and 20,000 followers. OpenAI’s findings underscore how Russian-linked actors are exploiting generative AI to amplify propaganda, echoing prior incidents of AI-generated election meddling and cyber-crime.