OpenAI has disclosed dismantling a furtive Iranian influence operation leveraging its ChatGPT chatbot to create content aimed at the upcoming U.S. presidential polls and other topics, according to CyberScoop.
Attackers also utilized ChatGPT to establish long-form articles and comments regarding Latinx rights in the U.S., the ongoing Israel-Gaza conflict, Israel's Olympic presence, Venezuelan politics, and Scottish independence, all of which have been spread across five fake news websites, said OpenAI. "We identified a dozen accounts on X and one on Instagram involved in this operation, with some accounts posing as progressives and others as conservatives. They generated some of these comments by asking our models to rewrite comments posted by other social media users," added OpenAI, which also noted the campaign's failure to obtain substantial audience engagement. Such a development comes just days after separate covert Iranian cyberattacks were reported by both Microsoft and Google's Threat Analysis Group to have targeted the U.S. presidential campaigns.
