OpenAI Disrupts High-Impact Russian and Iranian AI Influence Campaigns

Key points

• OpenAI identified a Russian-linked operation using a fake persona named Mia Clark to control the Social Research Center, a front for propaganda in Latin America.
• The Russian campaign generated fake audio clips and manipulated school events in Ecuador and Peru, leading to official denials from local governments.
• A separate Iranian operation created seven fake journalist personas to publish nearly 100 articles in small and medium outlets focused on international affairs.
• OpenAI rated the Russian campaign a 5 and the Iranian campaign a 4 on its impact scale, noting they achieved higher reach than typical low-impact AI campaigns.
• The Russian operators used VPNs to bypass regional restrictions and prompted the model in Russian, while Iranian operators used Persian and English prompts.

Background

This development follows earlier OpenAI actions in August 2026 against Russian-linked accounts promoting a fake 'Burke Sovereignty Index' and other disinformation. It also occurs amid shifting perceptions in Latin America, where recent polls indicate China is now viewed more positively than the US, creating a volatile environment for foreign influence operations. The use of AI for voice cloning and fake personas has also raised legal and safety concerns, as seen in recent lawsuits regarding ChatGPT's impact on user mental health.

How outlets are covering it

NBC News highlighted the operation's reach into schools and political figures in Latin America, emphasizing the disruption of the 'Dark Clark' network. CyberScoop detailed the specific tactics, including the use of fake audio clips in Ecuador and Bolivia, and noted the operators' interest in the Wagner successor group Politology. VSquare provided background on the International Burke Institute, a related fake think tank, and noted that while the immediate reach was modest, the infrastructure built was significant. Unite.AI focused on the technical details, such as the operators' use of VPNs and the specific impact ratings assigned by OpenAI. All sources agree that these campaigns represented a significant escalation in the sophistication of AI-assisted influence operations, moving beyond simple social media spam to planting content in legitimate media and academic contexts.

Why it matters

The disruption of these operations highlights the growing threat of AI-enabled influence campaigns that can bypass traditional media gatekeepers and reach mainstream audiences. The Russian campaign's success in generating official denials from governments in Latin America indicates a potential for real-world political impact, while the Iranian campaign's ability to plant articles in multiple outlets shows the scalability of AI-generated content. These events underscore the need for robust detection and mitigation strategies by tech platforms and governments to counter foreign influence operations that leverage artificial intelligence to create credible but false narratives.

What to watch

OpenAI has shared information on both cases with relevant authorities and industry partners. The company expects continued scrutiny of AI-generated content in media and academic contexts. Regulators and platform operators may increase efforts to detect and ban accounts associated with such influence operations, particularly those using fake personas or front organizations. The ongoing activity of related entities, such as the International Burke Institute, suggests that these operations may continue to evolve and adapt to new detection methods.