Seven fake journalists got nearly 100 ChatGPT-polished articles into overseas outlets; OpenAI banned the Iranian operation

In a threat report published October 8, OpenAI said it banned accounts behind two influence operations, one from Iran and one from Russia, that used ChatGPT. The Iranian operation used seven fake bylines posing as American and other Western journalists to pitch long-form articles about the US-Iran war to more than a dozen online outlets covering international affairs. ChatGPT revised drafts to fit each outlet's submission rules and wrote the pitch emails to editors. Nearly 100 articles were published or syndicated under those bylines.

The Russian operation set up a fake think tank in Latin America, the Social Research Center, staffed by locals who did not know they were working for a Russian group. It spread forged letters and fake audio in Peru, Ecuador and Bolivia, some of which drew official denials. OpenAI rated it Category 5 of 6 on the IO Breakout Scale, the first Category 5 among the 30 operations it has reported since early 2024. The Iranian operation reached Category 4.