OpenAI reports that it banned two covert influence operations, one from Russia and one from Iran, that used its models alongside traditional techniques to support sophisticated "false front" entities. The operations laundered geopolitical, conflict-related messaging into target audiences. The Iranian operation used seven "journalist" personas to pitch long-form articles to small and medium online outlets worldwide, and also generated batches of social media comments on the US-Iran war. The Russian operation co-opted unwitting people in Latin America to run a "think tank," and created fake "leaked" documents and audio scripts. Both made heavy use of AI to draft internal reports, often using deceitful methods to exaggerate their effectiveness.
The operations closely resembled pre-AI influence campaigns, such as the fake journalist "Alice Donovan" linked to Russian military intelligence and the "Peace Data" outlet run by individuals tied to the Russian Internet Research Agency. Using the IO Breakout Scale, OpenAI assessed the Russia-origin operation as Category 5, the first of that level it has disrupted, while the Iran-origin operation reached Category 4. Both managed to land content in mainstream media outlets rather than relying only on social media.
OpenAI notes that operations placing content in real media tend to have the highest reach and impact, based on its review of 30 covert influence operations exposed since early 2024. AI can add scale, efficiency, linguistic fluency and editorial polish to such campaigns. Operators can exploit employees and editors who have no idea who was behind the content or what its motivations were.
Source: OpenAI Blog · Summarized by HeadlinesBriefing