OpenAI bans Russian ChatGPT accounts used in covert misinformation campaign

OpenAI has announced the shutdown of numerous ChatGPT accounts linked to Russia following the discovery of a covert influence campaign. According to reporting from CNBC and Anadolu Ajansı, these accounts were systematically banned after being identified as tools for a misinformation effort. The operation involved the use of AI to create pro-Moscow narratives and distribute them across various social media platforms. OpenAI confirmed the disruption of this activity in its own official statement, describing the effort as a new covert influence campaign originating from Russia. Coverage from Yahoo and Le Monde.fr emphasizes the sophisticated nature of the deception used by the actors involved.

The campaign reportedly utilized ChatGPT to build a sense of fake expertise, which included the creation of fake Telegram channels and the production of fake videos. Le Monde.fr specifically highlights the establishment of a real-fake Israeli think tank as part of this strategic effort to mislead audiences. These details underscore the scale of the operation, which sought to blend AI-generated content with fabricated institutions to lend credibility to its messaging. This development is significant because it demonstrates the evolving methods of state-linked disinformation actors. The Register characterizes the operation as a process of using a "slop factory" to create more "slop," referring to the mass production of low-quality AI content.

By leveraging large language models, the Russian actors were able to automate the generation of narratives designed to influence public perception on a global scale. The use of AI to mimic professional think tanks and authoritative sources represents a tactical shift toward creating high-fidelity fake identities to bypass traditional skepticism. Future developments will likely center on how OpenAI and other AI providers detect similar covert operations. The current coverage establishes that OpenAI has already taken action by banning the identified accounts, but it does not specify the exact number of accounts closed or the total reach of the fake Telegram channels. Observers will be watching for further disclosures from OpenAI regarding the technical signatures used to identify these Russia-linked accounts and whether other platforms have detected similar patterns of AI-driven influence campaigns targeting social media users.