The Iranian network that deceived media worldwide is exposed | Report

OpenAI has uncovered an Iranian influence operation that used ChatGPT to generate approximately 100 fake articles across 12 global media outlets, employing seven fictitious journalists. The articles criticized US and Israeli policy toward Iran. Following publication, fake comments were generated on social media. OpenAI blocked the accounts involved but did not definitively identify the operating entity.

OpenAI has revealed a broad Iranian influence operation that used artificial intelligence to inject approximately 100 fake articles into 12 media outlets worldwide. The operation relied on seven fictitious journalists, under whose names articles were written criticizing the American and Israeli war on Iran. The operators used ChatGPT in Persian to improve phrasing and tailor publication pitches. Among the identified headlines were: "Is US policy toward Iran moving into private hands? The Jared Kushner question" and "The war on Iran and the collapse of the West's moral credibility." The publishing outlets are not among the largest in the US, but some have significant social media followings, such as Middle East Monitor with nearly two million Facebook followers. After publication, the operators asked ChatGPT to generate fake social media comments portraying the US as an aggressor, Israel as an unreliable ally, and Iran as a victim. OpenAI blocked the accounts involved but emphasized it did not definitively identify the operating entity, noting it could be a commercial actor running a paid influence campaign rather than a direct operation by the Iranian regime.