OpenAI bans Russian, Iranian "false front" operations
OpenAI says it has banned Russian and Iranian influence operations that used its models to run fake journalists and a front think tank, including the first Category 5 operation it has disrupted.
OpenAI has banned account clusters tied to two covert influence operations, one from Russia and one from Iran. According to its latest threat report, both used its models alongside traditional techniques to run "false front" entities that pushed geopolitical and conflict-related messaging to target audiences.
The Iranian operation ran seven fake "journalist" personas that pitched long-form articles to small and mid-sized online outlets around the world. It also generated batches of social media comments, mostly about the US-Iran war. OpenAI rated it Category 4 on the six-point IO Breakout Scale.
The Russian operation, which OpenAI nicknamed "Dark Clark," targeted Latin America. It appears to have controlled a self-described research platform, the Social Research Center, through a fake persona called "Mia Clark," and to have drawn in unwitting local people to run it. Much of the activity aimed to damage Ukraine's reputation in the region, while some targeted political outcomes in Argentina and Bolivia. The operators also produced fake "leaked" documents and audio scripts. They reached OpenAI's services through VPNs, since the company blocks access from Russia, and OpenAI says it has shared details with authorities. It rated the operation Category 5, the first at that level it has disrupted since it began reporting.
Both operations got content into mainstream media outlets, though OpenAI notes that not all of it came from its models. Both also used AI mostly to draft internal reports, often exaggerating their own impact.
OpenAI's main conclusion is that the operations closely resembled pre-AI campaigns such as the fake journalist "Alice Donovan" and the "PeaceData" outlet. In its account, AI made the workflows easier but did not change the playbook. Across 30 operations it has exposed since early 2024, the ones that placed content in real outlets had the highest reach.