Full Breakdown
OpenAI Disrupts Russian and Iranian AI-Powered Influence Campaigns
By Drooid · · How we work
Core Event: Shutdown of ChatGPT-Based Propaganda Networks
On Oct 8, OpenAI announced it had terminated two influence operations that used its ChatGPT model to generate false journalist personas and support covert “think-tank” fronts. The accounts originated in Russia and Iran and were employed to produce and disseminate disinformation about the Ukraine conflict and the U.S.–Iran war. OpenAI’s blog detailed that the banned accounts created internal memos, long-form articles, and social-media comments designed to appear authentic before being placed in online outlets.
Background & Context
The Russian campaign focused on Latin America, where operatives impersonated regional institutions—such as a Peruvian educational authority—to organize school activities that highlighted controversial Ukrainian nationalist Stepan Bandera. The effort aimed to undermine support for Ukraine and sway local politics.
The Iranian effort targeted English- and Persian-language audiences, pushing anti-U.S. narratives about the war in Iran. Operatives used ChatGPT to draft articles, editorial pitches, and comment-thread replies that portrayed the United States as the aggressor and Iran as a resilient victim. Both campaigns leveraged AI to reduce the labor required for content creation while masking the operators’ identities.
Data & Statistics
- Seven fabricated journalist profiles authored articles.
- OpenAI identified almost 100 articles published or syndicated across roughly a dozen outlets worldwide.
- The earliest identified article appeared in 2025, with the latest in 2026; posting frequency rose after the U.S.–Iran war began in 2026.
- Russian operatives also ran a fake think-tank called the Social Research Center (SRC) in Latin America, recruiting unwitting researchers to produce reports.
Official Statements & Responses
OpenAI emphasized that AI-generated content was a minority of the workload for the Russian actors, who primarily used ChatGPT to update superiors. The company blocked the offending accounts and is monitoring for similar misuse.
On-the-Ground Reports
Peruvian outlet *Extra* confirmed that a school event featuring Bandera took place after the Russian operatives impersonated a regional education authority. A far-right Polish member of the European Parliament subsequently called for bans on individuals expressing support for Bandera, illustrating how the fabricated story fed existing historical tensions.
Why It Matters
The incidents show how readily available AI tools can be weaponized to produce sophisticated disinformation at scale, allowing foreign adversaries to “hide their hand” and launder content through legitimate-looking channels.
Conflicting Reports & Gaps
OpenAI could not identify the commercial client behind the Iranian campaign, leaving direct state sponsorship unresolved. While the company linked the Russian operation to a “think-tank” front, it did not confirm whether the SRC was formally controlled by a Russian intelligence entity.
Verbatim Quotes
- “The Russian fake thus both fed on, and fed into, historical tensions between Ukrainians and Poles,” — OpenAI
- “It’s remarkable that the Iranians managed to land their content in real media outlets, expanding the potential reach of their campaign,” — Jessica Brandt
- “We identified almost 100 articles published or syndicated under the operation’s bylines across roughly a dozen online outlets around the world,” — OpenAI