FAKE JOURNALISTS
The Iranian operation in the meantime concerned using ChatGPT to refine articles concerning the US-Iran conflict and to subsequently draft electronic mail pitches to actual media shops, mentioned OpenAI, suggesting the exercise appeared in line with that of a employed business actor.
OpenAI recognized practically 100 articles that had been printed or syndicated to a couple of dozen shops between July 2025 and October 2026. These had been written beneath seven fictitious bylines, utilizing the personas of American or Western journalists masking international affairs and geopolitics.
Some personas had accompanying social media accounts to create the impression of legitimacy, however in a minimum of one case, platform transparency settings confirmed the accounts had been primarily based in Iran.
Some articles had been additionally promoted by the social media accounts of the media shops, “giving the Iranian operators a possible approach to entry audiences that they might not in any other case have reached”, added OpenAI, which rated this side of the marketing campaign a Class 4 on the affect operations scale.
A much less profitable strand of the marketing campaign concerned utilizing ChatGPT to generate social media feedback on the US-Iran battle in each English and Persian. OpenAI mentioned the feedback it discovered usually drew likes, views and replies in solely single or double digits, with many coming from the operation’s personal accounts.
“In truth, the operators’ inner stories used a misleading calculation to measure their social media affect, presumably in an try and inflate the operators’ obvious effectiveness,” it added.
OpenAI has periodically reported on affect operations it has detected up to now. In June 2026, it mentioned it disrupted two Chinese language-linked operations, one which ran a knowledge centre-focused marketing campaign and the opposite geared toward criticising US tariffs.
