OpenAI Suspends Accounts in Russian and Iranian Influence Operations
OpenAI's own rating of the two operations at 4 to 5 on its six-point scale puts AI-assisted false-front campaigns at the high end of the reach it tracks, with the Dark Clark material spreading far enough to draw an Ecuadorian minister's denial and a newspaper fact-check.
Reporting from 1 source: GIGAZINE.
OpenAI says it suspended accounts tied to two influence operations that paired conventional tradecraft with AI-generated people and organizations. The Russian operation, Dark Clark, ran a fictitious research platform called the Social Research Center and pushed disinformation about Ukraine and local leaders in Argentina and Bolivia. The Iranian operation, False Signature, generated articles under seven bylines, and nearly 100 pieces were published or republished by more than a dozen outlets worldwide.
The Iranian operation got its name from seven bylines: the threat actors prompted in Persian and produced Persian and English text, including social media comments, pitch drafts, and long-form articles. Nearly 100 of those articles ran in more than a dozen outlets, though social media reaction to them was limited. The Russian operation built its cover differently, using a fictitious person named Mia Clark to run a self-described research platform whose activity reports lent the work plausibility. Recruits in Latin America worked for that platform without knowing who was behind it, and Ecuador's minister of education issued a statement denying the disinformation it spread. OpenAI notes that the likelihood of spread depends heavily on the distribution channels involved.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.