Iran and Russia used ChatGPT for propaganda by ‘fake journalists’, OpenAI says

In a recent development, OpenAI disclosed that it had terminated two influence campaigns, one originating from Russia and the other from Iran, which utilized ChatGPT to impersonate journalists and propagate anti-Ukraine sentiments, with the aim of inciting discord in the US and Israel over the Iran conflict.

The Russian operation entailed the recruitment of unsuspecting individuals in Latin America to work for a think-tank named the Social Research Center (SRC), a covert entity designed to undermine Ukraine's image in the region and sway local politics. OpenAI revealed that the SRC's staff in Latin America were largely unaware that they were working for a Russian entity.

The operation leveraged ChatGPT to draft internal reports, generate content, and compose social media posts and comments, often targeting divisive topics such as Polish far-right MEP Grzegorz Braun's proposal to declare anti-Polish individuals persona non grata in Poland.

Simultaneously, the Iranian operation employed seven fabricated journalist profiles, all claiming to be from Western nations including the United States, to craft long-form articles and disseminate them through various international media outlets. These articles, published between July 2025 and October 2026, predominantly appeared in small to medium-sized publications focusing on global affairs, geopolitics, and Middle Eastern events.

OpenAI identified nearly 100 articles linked to the operation across roughly a dozen online platforms worldwide.