OpenAI says it banned Russian and Iranian ChatGPT influence operations
OpenAI said on Thursday it shut down two influence operations, one from Russia and one from Iran, that used ChatGPT to support "false front" entities including fake journalists and a purported think tank.
1 / 3
The story, neutrally told
Mixed · 2OpenAI said on Thursday it banned a cluster of ChatGPT accounts that originated in Russia and Iran and were used to support "false front" entities for geopolitical and conflict-related messaging. France 24N “The American AI giant said it banned a cluster of ChatGPT accounts that originated in Russia and Iran and were used to support "false front" entities for geopolitical and conflict-related messaging.” Read at France 24 ↗ Washington ExaminerR “OpenAI banned numerous ChatGPT accounts that originated in Russia and Iran.” Read at Washington Examiner ↗ Mixed · 2The Iranian operation, which OpenAI dubbed "Bogus Bylines", used seven fake journalist personas, some purporting to be from Western countries including the United States, to pitch and plant long-form articles in news outlets; it also generated social media comments on the US-Iran war. France 24N “The Iranian operation involved seven "journalist" profiles, each purporting to be from Western countries including the United States, that were used to plant long-form articles in media outlets around the world.” Read at France 24 ↗ Washington ExaminerR “The operation, which the AI firm dubbed “Bogus Bylines,” also generated social media comments on the geopolitical conflict.” Read at Washington Examiner ↗ Centre · 1OpenAI said it identified almost 100 articles published or syndicated under the operation's bylines across roughly a dozen small to medium online outlets focused on international affairs, geopolitics and the Middle East. The earliest dated from July 2025 and the latest from October 2026, with publication increasing after the US-Iran war began earlier this year. France 24N “"We identified almost 100 articles published or syndicated under the operation's bylines across roughly a dozen online outlets around the world," OpenAI said.”“The earliest article identified by OpenAI was published in July 2025, and the latest in October 2026, with the frequency of reports increasing after the US-Iran war broke out earlier this year.” Read at France 24 ↗
Centre · 1One persona, Ervin B. Hoskins, described as "an American freelance writer", had accounts on X and Instagram. According to screenshots OpenAI provided, X's transparency information linked the account to a "West Asia android app" and Instagram's showed it was based in Iran; both accounts appeared to be suspended. France 24N “X's transparency information showed the account was connected via a "West Asia android app" and Instagram's transparency information showed the account was based in Iran, according to screenshots provided by OpenAI.” Read at France 24 ↗ Mixed · 2The Russian operation ran a supposed think tank, the Social Research Center (SRC), as a front in Latin America, aimed at undermining Ukraine's reputation in the region and influencing local political outcomes. It also created fake leaked documents and audio scripts. France 24N “The Russian operation co-opted "unwitting people" in Latin America to run a "think-tank" -- called the Social Research Center (SRC)” Read at France 24 ↗ Washington ExaminerR “Those behind the Russian operation ran a supposed think tank as a front organization in Latin America and created fake leaked documents and audio scripts.” Read at Washington Examiner ↗ Centre · 1OpenAI said the available evidence indicates the SRC's Latin American employees did not know they were working for a Russian group, and that the SRC produced a majority of original content through its co-opted staff. It called this the most complex attempt to run a front identity it had disrupted over the past two and a half years. France 24N “"The available evidence indicates that the SRC's employees in Latin America were not aware that they were working for a Russian group," OpenAI said.”“This stands out as the most complex attempt to run a front identity that we've disrupted over the past two and a half years.” Read at France 24 ↗
Right · 1Both operations, particularly the Russian one, used the chatbot to draft internal reports. OpenAI said the operations' goal was to "launder geopolitical, conflict-related messaging into their target audiences", and that disclosing them should make further research and disruption easier and continuing the operations harder. Washington ExaminerR “Both operations, particularly the one in Russia, made use of the chatbot to draft internal reports.”“OpenAI explained that by reporting the “false front” operations, it hopes to “make further research and disruption easier — and make continuing the operations harder.”” Read at Washington Examiner ↗ Mixed · 2The Washington Examiner placed the disclosure in a safety report context and noted that rival Anthropic has released similar threat intelligence reports, while France 24 presented it as an illustration of how state actors can exploit widely available AI tools at scale. Washington ExaminerR “Like its competitor, Anthropic has released similar safety reports.” Read at Washington Examiner ↗ France 24N “OpenAI's report is the latest illustration of how state actors can easily exploit widely available AI tools to peddle sophisticated propaganda against adversaries on a mass scale.” Read at France 24 ↗
Every sentence links to the reporting it rests on. The pill in front of each says where its sources sit: Left, Centre or Right when one side supplies at least half of them, Mixed when they are evenly split. The number is how many outlets it cites.
Left1 outlet
No left outlet in our sources has covered this story yet.
Centre1 outlet
- Framing
- Frames the story as a case of state actors exploiting widely available AI tools for propaganda at scale, leading with the Russian and Iranian networks.
- Emphasis
- Operational detail: seven Iranian personas, almost 100 articles, the Ervin B. Hoskins persona, and the Russian Social Research Center using unwitting Latin American staff.
- Leaves out or plays down
- Does not mention the internal-report drafting or the Anthropic comparison.
- Charged language
- “propaganda networks”“peddle sophisticated propaganda”
- For example
-
“OpenAI's report is the latest illustration of how state actors can easily exploit widely available AI tools to peddle sophisticated propaganda against adversaries on a mass scale.” — France 24
Right2 outlets
- Framing
- Presents it as OpenAI cracking down on foreign threat actors, within a wider context of AI safety reporting, public concern and industry transparency.
- Emphasis
- The 'Bogus Bylines' name, fake leaked documents and audio scripts, OpenAI's stated aim of making disruption easier, and comparison with Anthropic's reports.
- Leaves out or plays down
- Leaves out the article count, dates, the Hoskins persona and the Social Research Center name, and OpenAI's remark that the SRC staff were unwitting.
- Charged language
- “cracked down”“threat actors”“foreign actors”
- For example
-
“OpenAI cracked down on threat actors in Russia and Iran that made use of the company’s chatbot to support “false front” entities designed to carry out covert influence operations around the world.” — Washington Examiner
What every side reports
- OpenAI said on Thursday it banned ChatGPT accounts originating in Russia and Iran.
- The Iranian operation used seven fake journalist personas to place articles and generate social media comments about the US-Iran war.
- The Russian operation ran a purported think tank in Latin America aimed at undermining Ukraine's reputation and influencing local politics.
OpenAI organisation
Says it banned the accounts and published details of the operations to make further research and disruption easier and continuing the operations harder; calls the Russian think-tank front the most complex it has disrupted in two and a half years.
“OpenAI explained that by reporting the “false front” operations, it hopes to “make further research and disruption easier — and make continuing the operations harder.”” — Washington Examiner
“This stands out as the most complex attempt to run a front identity that we've disrupted over the past two and a half years.” — France 24
Russia place
Per OpenAI, Russia-origin actors ran a front think tank in Latin America against Ukraine's reputation. No Russian response is reported in the coverage.
“The foreign actors used ChatGPT to undermine “Ukraine’s reputation in the region” or influence “local political outcomes,” OpenAI said.” — Washington Examiner
Iran place
Per OpenAI, Iran-origin actors used fake Western journalist personas to place articles about the US-Iran war. No Iranian response is reported in the coverage.
“The Iranian operation involved seven "journalist" profiles, each purporting to be from Western countries including the United States” — France 24
Centre1 article
-
OpenAI bans Russian, Iranian ChatGPT propaganda networks
Neutral Detailed, OpenAI-sourced account of the two networks, framed as evidence of how easily state actors misuse AI tools.
Right2 articles
-
OpenAI bans ‘false front’ influence operations in Russia and Iran
Neutral Shorter account of OpenAI's crackdown on foreign threat actors, tied to industry safety reporting and Anthropic's similar reports.
-
- 9 Oct 00:42 First France 24N OpenAI bans Russian, Iranian ChatGPT propaganda networks
- 9 Oct 00:53 +11m Washington ExaminerR OpenAI bans ‘false front’ influence operations in Russia and Iran
- 9 Oct 01:08 +25m The Straits TimesRC OpenAI says Iran, Russia influence ops busted via AFP
- 9 Oct 01:25 +43m NPRLC OpenAI caught Russians and Iranians using ChatGPT for influence campaigns
Times are when each article was published, or when we first saw it if the outlet gave no time.