ERBIL, Kurdistan Region - OpenAI on Thursday banned a cluster of ChatGPT accounts originating in Iran it said used “sophisticated false front entities” posing as Western journalists to push state-backed political agendas.
“The Iranian operation included a stable of seven ‘journalist’ personas which it used to pitch long-form articles to small and medium online outlets around the world,” including generating batches of social media comments “on topics related to the US-Iran war,” to “launder geopolitical, conflict-related messaging into their target audiences.”
The operation “made heavy use of AI to draft internal reports” and “used questionable or outright deceitful methodologies to exaggerate the operators’ effectiveness,” according to the report, which noted its unusual potential reach of achieving an influence operations (IO) Category four breakout scale and using various methods of landing their content - not all of which was generated from Open AI models - in mainstream media outlets.
The report illustrates challenges underpinning the modern digital landscape and susceptibility to the wider exploitation and weaponization of artificial intelligence tools in peddling disinformation campaigns on a mass scale.
Established in 2015 in San Francisco, California, OpenAI measures a breakout scale impact using six categories. Category Four operations transcend from social media into mainstream media, while Category Five operations are amplified by high-profile individuals such as celebrities and political candidates. An IO reaches Category Six if it triggers a policy response or concrete action, including calls for violence.
The earliest article identified was published in July 2025, and the latest in October 2026, the report stated, noting an increased “density of publications on US-Iran relations from March 2026 onwards,” after the outbreak of the US-Iran conflict.
The tech firm identified almost 100 articles published or syndicated across roughly a dozen small to medium outlets around the world under bylines from “Western journalists” writing on international politics, the Middle East, human rights, or conflict: Ervin B. Hoskins, Noah Lamington, Sophia Gonzalez, Michael Harrison, Ericka Feusier, Jenny Williams, and Alice Johnson. The report noted that Meta in March disrupted an Iranian IO under the Michael Harrison byline.
The report identified the X account for the Ervin Hoskins persona showing the use of the West Asia Android app and the Instagram account transparency information showing location in Iran. Meta banned the Instagram account in August.
According to the report, the journalist personas succeeded in getting their articles published by a range of outlets, some with substantial followings, including one with almost 2 million followers on Facebook, almost 355,000 followers on X, and over 544,000 on Instagram as of August 2026.
OpenAI has documented and curtailed Iran’s use of its digital infrastructure in past instances.
OpenAI in August 2024 shut down accounts of an Iranian group for using its ChatGPT chatbot to generate content directed at American voters ahead of the US presidential election and other issues, according to reporting from the Guardian, by generating long-form articles and shorter social media comments on topics relating to Gaza and Israel, which at the time “did not appear to have achieved meaningful audience engagement,” the tech firm said.
