OpenAI has banned accounts linked to a previously unreported operation dubbed "No Bell", likely originating from Russia, which used AI to produce critical content targeting the US and its allies for African audiences. The operation involved a ChatGPT account generating long-form articles and social media posts about geopolitics in sub-Saharan Africa. The user primarily prompted in English but sometimes used Russian-language instructions attributed to a manager.

Content was posted on Facebook pages posing as news outlets in South Africa, Ghana, Kenya, and Angola. Meta subsequently banned these pages. The operation attempted to conceal AI-generated text by removing em-dashes and mimicking human journalistic style.

A fabricated persona, "Dr. Manuel Godsin," purportedly a PhD holder from the University of Bergen, was used as a byline for articles published on African news websites. OpenAI found no credible evidence of this person's existence, linking the photo to a St. Petersburg law student. This case was originally published in OpenAI's February 2026 report on disrupting malicious AI uses.

Coordinated efforts with Meta identified the network's reach and tactics, highlighting the evolving challenges of AI misuse in geopolitical influence campaigns.