OpenAI reports Russian network used ChatGPT for influence campaign

A network originating in Russia used ChatGPT to promote an organization that presented itself as an academic institute, according to an OpenAI report published Wednesday.

OpenAI said it banned a group of accounts whose operators used Russian-language prompts to generate social media posts and comments, mostly in English. They also instructed ChatGPT to conceal signs that the content originated in Russia.

“Although the campaign appears to have reached relatively small audiences, its elaborate construction distinguishes it from other Russia-linked influence operations we have disrupted since the start of the war in Ukraine,” the company wrote.

Because ChatGPT is unavailable in Russia, OpenAI said the group used virtual private networks (VPNs) to disguise their location and access the service.

The group used ChatGPT to promote the International Burke Institute, which claimed to be an Israel-based expert group, through official-looking profiles and accounts posing as regular users. OpenAI found that 34 of 36 articles it reviewed on the institute’s website were copied, and some were credited to academics who did not write them.

“Some of these articles were years old; others were attributed to the wrong authors,” OpenAI wrote. “For example, one article on the China-Pakistan Economic Corridor appears to have been copied from a Cambridge University Press original, but incorrectly attributed to a professor at the University of Nottingham whose expertise is in South Asian politics.”

Similar political influence campaigns have also used generative AI.

In 2024, OpenAI identified a separate Russia-linked campaign called Doppelganger. The network used ChatGPT to create articles, social media posts, and comments that praised Russia and attacked Ukraine, NATO, and the United States.

Russia-linked groups have also used ChatGPT for cyber operations. OpenAI said Forest Blizzard, a group tied to Russian military intelligence, used the chatbot to research targets, write scripts, and support phishing attacks.

While the group may have used Russian language prompts, OpenAI did not attribute the IBI campaign to the Russian government.

According to OpenAI, some people in Israel may have represented the institute at conferences or through article submissions. Still, their relationship to the campaign remains unclear.

“We are not in a position to determine the relationship between these individuals, the IBI, and the operators in Russia,” OpenAI wrote.