OpenAI Exposes Russia and Iran Using ChatGPT for Covert Influence Operations and Fake News

OpenAI Disrupts Sophisticated State-Sponsored Influence Operations

OpenAI announced Thursday it banned ChatGPT accounts linked to Russia and Iran that were running elaborate influence operations targeting multiple countries1

1

2

2

. The AI company revealed that these state-sponsored influence operations used ChatGPT alongside traditional techniques to create and operate sophisticated "false front" entities for geopolitical messaging4

4

. According to OpenAI, the Russian campaign represents the widest-reaching influence operation the company has disrupted in over two and a half years of monitoring such activity5

5

.

The findings underscore how easily state actors can exploit widely available AI tools to peddle propaganda on a mass scale2

2

. Both operations made heavy use of AI to draft internal reports, with Russian operatives primarily using ChatGPT to update their superiors about campaign progress4

4

.

Iranian Operation Created Fake Journalist Personas Across Global Media

The Iranian operation, dubbed "Bogus Bylines" by OpenAI, deployed seven fake journalist personas claiming to be from Western countries including the United States3

3

. These AI-driven propaganda networks published nearly 100 articles across roughly a dozen small to medium online outlets worldwide, generally focused on international affairs, geopolitics, and events in the Middle East2

2

.

One persona named Ervin B. Hoskins presented himself as "an American freelance writer" with accounts on X and Instagram3

3

. However, transparency settings revealed the accounts were actually based in Iran, with X showing connection via a "West Asia android app"5

5

. Both accounts have since been suspended.

The earliest article identified by OpenAI was published in July 2025, with the latest in October 20263

3

. The frequency of reports increased significantly after the U.S.-Iran war broke out earlier this year. The operation also generated social media comments on topics related to the conflict in both English and Persian4

4

.

Publications that fell victim to these AI-driven disinformation campaigns included the progressive U.S. news site Daily Kos and the U.K.-based Middle East Monitor, which published at least 21 articles by four fake authors before removing the author pages1

1

.

Russian Covert Think-Tank Spread Anti-Ukraine Propaganda in Latin America

The Russian operation, referred to as "Dark Clark," proved even more sophisticated by co-opting unwitting people in Latin America to run a fake think-tank called the Social Research Center (SRC)5

5

. This covert think-tank was used for activities aimed at undermining Ukraine's reputation in the region and influencing local politics2

2

.

"The available evidence indicates that the SRC's employees in Latin America were not aware that they were working for a Russian group," OpenAI stated3

3

. The operation recruited unwitting researchers to interview experts and compose reports, similar to a 2020 Russian operation linked to Yevgeny Prigozhin's Internet Research Agency1

1

.

The Russian operatives used ChatGPT for influence campaigns to write reports detailing their work to higher-ups1

1

. One campaign impersonated a regional educational authority in Peru, tricking schools into organizing activities about Ukraine featuring Stepan Bandera, a controversial figure admired by nationalists but reviled by others for his movement's collaboration with Nazis1

1

. A Polish news report about this event prompted a far-right Polish member of the European Parliament to suggest banning entry of people supporting Bandera. "The Russian fake thus both fed on, and fed into, historical tensions between Ukrainians and Poles," OpenAI wrote1

1

.

Darren Linvill, co-director of Clemson University's Watt Family Innovation Center Media Forensics Hub, noted that "Russians are constantly engaging in these sorts of active measures" involving narrative laundering and fake narratives aligned with Russia's political agenda1

1

.

Implications for AI Governance and Platform Security

OpenAI's investigation found that campaigns focused on placing fake narratives in established media outlets reached wider audiences than those relying solely on fake social media accounts1

1

. While Russian operatives did use AI to generate fake material, it represented a minority of their workload1

1

.

"These Russian underlings were trying to use AI to create a shortcut and do less work in the same way that a lot of us use AI, but in this case it helped get them caught," Linvill explained1

1

.

The disclosure arrives as American public trust in AI and lawmakers' ability to handle its development continues to decline. A recent Reuters/Ipsos survey found 57 percent of respondents believe the Trump administration isn't taking AI risks seriously enough, while 54 percent said the same of Congress4

4

. The administration has favored voluntary safeguards over federal regulations, inviting tech leadership to agree to "morally-binding" self-policing pledges, though critics question whether companies can effectively self-regulate4

4

.

OpenAI stated that reporting on these AI-driven disinformation campaigns will make "further research and disruption" easier and has in some cases deterred similar activity4

4

. As geopolitical influence operations grow more sophisticated, the tension between AI innovation and security safeguards will likely intensify, raising questions about how platforms can detect and prevent state-sponsored propaganda without stifling legitimate use.