OpenAI says Iran, Russia influence ops busted

OpenAI announced the shutdown of two influence operations, one originating from Russia and another from Iran, that utilized ChatGPT to disseminate fake news and anti-Ukraine propaganda. The report highlights how state actors can exploit readily accessible AI tools to carry out sophisticated propaganda campaigns on a large scale.

OpenAI identified and banned a series of ChatGPT accounts tied to Russia and Iran, which were used to create fake journalist personas for geopolitical messaging. In the Iranian operation, seven profiles were created, each claiming to be from Western nations like the United States. These accounts were used to publish numerous articles in various online outlets, primarily covering international affairs and Middle Eastern events.

The operation began in July 2025 and peaked after the US-Iran conflict started this year, with articles continuing to appear until October 2026.

The Russian operation involved a more complex scheme, where unwitting individuals in Latin America were recruited to run a think-tank called the Social Research Center (SRC). The SRC's employees were unaware they were working for a Russian group. The think-tank's content was largely generated by co-opted staff, making it the most intricate operation disrupted by OpenAI in the past two and a half years.