OpenAI said on Thursday it had shut down two state-linked influence operations, one from Russia and one from Iran, that used ChatGPT to run fake journalist personas and a front "think-tank" pushing geopolitical messaging. The Iranian network used seven fake Western "journalists" to place almost 100 articles in about a dozen small and mid-sized outlets, with output rising after the US-Iran war began this year, while the Russian one used unwitting staff in Latin America at a so-called Social Research Center to undermine Ukraine's reputation, which OpenAI called the most complex front-identity effort it has disrupted in two and a half years. The findings show how easily widely available AI tools can be used to scale up covert propaganda.