OpenAI blocks Russian and Iranian influence operations using ChatGPT

OpenAI reported that it had blocked two influence operations—one Russian and one Iranian—that used its language models together with traditional techniques to run sophisticated front organizations.

The Russian operation employed ChatGPT accounts to target audiences in Latin America, seeking to damage Ukraine’s reputation and affect political outcomes in Argentina and Bolivia, and involved unaware local participants managing a “study centre.” The Iranian operation created seven fake journalist identities to pitch lengthy articles to small and medium online outlets worldwide.

OpenAI shared information about the Russian operation with competent authorities.