OpenAI dismantles Iranian, Russian influence campaigns on platform

OpenAI has taken action against two separate influence operations originating from Iran and Russia that weaponised its ChatGPT platform to conduct coordinated propaganda campaigns. The company identified and disabled accounts involved in creating fabricated personas and content designed to deceive audiences and manipulate public discourse. According to OpenAI's findings, the bad actors established fake journalist identities and pseudo-research organisations to lend credibility to their messaging efforts. The campaigns systematically generated anti-Ukraine narratives and disseminated them across multiple platforms, exploiting the reach and persuasiveness of AI-generated content to amplify geopolitical messaging. OpenAI said it has shared technical indicators and additional details of the operations with relevant authorities and platforms to prevent similar activities. The disclosure underscores growing concerns about the misuse of generative AI tools by state and state-linked actors seeking to conduct sophisticated influence operations at scale. Major technology platforms have increasingly grappled with detecting and countering such coordinated inauthentic behaviour, particularly involving AI-generated content designed to evade traditional detection methods.