OpenAI Disrupts Russia and Iran Influence Campaigns on ChatGPT

OpenAI confirmed on Thursday that it had thwarted two information influence campaigns tied to Russia and Iran. The company stated that these operations utilized the ChatGPT platform to generate fictitious journalistic personas and operate a "research center" designed to covertly promote anti-Ukraine propaganda.

Iranian Network Details

The American firm reported banning a cluster of ChatGPT accounts originating from Russia and Iran, which were employed to support shell entities dedicated to advancing messages related to geopolitical issues and conflicts. This disclosure was cited by Agence France-Presse.

The Iranian component of the scheme involved seven individuals posing as "journalists" who claimed affiliations with Western nations, including the United States. These fabricated profiles were leveraged to publish lengthy articles across various international media outlets.

According to OpenAI, investigators identified nearly 100 articles published or republished under these names on approximately 12 electronic media platforms worldwide. The company noted that these platforms were generally small to mid-sized and focused primarily on international affairs and Middle East events.

Records indicate the earliest detected article appeared in July 2025, while the most recent publication dates back to October 2026. The operation also generated social media comments addressing topics connected to the US-Iran war.

One prominent figure in this network, Ervin B. Hoskins, presented himself as an "independent American writer" and maintained accounts on multiple platforms, including X and Instagram.

Transparency data from X revealed that the account was connected via an application operating from West Asia. Conversely, Instagram data suggested the account was based in Iran. Both accounts appear to have been suspended subsequently.

Russian Operations in Latin America

The Russian campaign relied on individuals in Latin America, acting without their knowledge, to manage an entity labeled the "Social Research Center." This structure facilitated covert activities aimed at undermining Ukraine's image in the region and influencing local policies.

OpenAI stated that available evidence indicates staff members of the center in Latin America were unaware they were working for a Russian group. The company added that the center depended on exploited employees to produce the majority of its original content.

Describing the disruption, OpenAI characterized the intervention as the most complex attempt to establish a fake front it has managed to stop during the past two and a half years.