Disrupting AI-enabled “false front” operations
OpenAI reports two AI-enabled covert influence operations from Russia and Iran that used fake personas and entities to spread geopolitical messaging, with one reaching the highest threat level on its scale.
Over the past two and a half years, OpenAI has documented how threat actors misuse AI models for cyber attacks, covert influence operations, scams, and policy violations. The company publishes these findings to inform regulators, peers, and the public about emerging AI-driven threats. Recent cases include two sophisticated operations—one from Russia and one from Iran—that combined AI with traditional tactics to create deceptive "false front" entities. These operations aimed to launder conflict-related messaging into target audiences through mainstream media outlets rather than social media alone.
The Iranian operation deployed seven fake journalist personas to pitch long-form articles to small and medium online outlets worldwide, while also generating social media comments on the US-Iran war. The Russian operation co-opted unwitting individuals in Latin America to run a "think tank" that produced fake leaked documents and audio scripts. Both operations used AI to draft internal reports and exaggerated their effectiveness through deceitful methodologies, closely resembling pre-AI influence tactics but with enhanced efficiency and scale.
OpenAI assesses the Russian operation as a Category 5 on its IO Breakout Scale—the highest threat level—and the Iranian operation as Category 4. These ratings reflect the operations' potential reach and impact, with both successfully placing content in mainstream media outlets. The pattern across 30 covert influence operations exposed by OpenAI shows that operations targeting real media outlets tend to achieve greater penetration than those relying solely on social media distribution.
AI enables false-front operations to scale more efficiently, with improved linguistic fluency and editorial capabilities that exploit unsuspecting victims such as editors or employees. OpenAI emphasizes that responsible disclosure remains a critical tool in disrupting such operations, as past exposed fronts like "Alice Donovan" and "PeaceData" ceased activity after being revealed. The company aims to make further research and disruption easier while increasing the difficulty for threat actors to continue their operations.