On Tuesday, the United States and its allies announced the successful seizure of a complex Russian propaganda network that relied on artificial intelligence (AI) technology to operate nearly a thousand covert accounts on the social media platform X. While governments have increasingly utilized AI in disseminating messages more effectively, this takedown is notable as Western intelligence agencies managed to trace the operation back to an officer of the Russian FSB intelligence agency and a former senior editor at the state-controlled publication RT (previously known as Russia Today). In a detailed joint advisory, agencies in the United States, Netherlands, and Canada identified various software programs utilized by the network, including the program Meliorator, which generated fictitious user accounts known as “souls" from different countries. The FBI obtained a court order enabling them to seize two web domains associated with the operation's registered email addresses. FBI Director Christopher A. Wray highlighted the significance of the enforcement action, stating, "Today's actions mark the disruption of a Russian-sponsored social media bot farm enhanced by Generative AI, which Russia intended to utilize for disseminating AI-generated foreign disinformation.

Their aim was to scale their efforts using AI to undermine our partners in Ukraine and manipulate geopolitical narratives advantageous to the Russian government. " While some automated accounts with more detailed profiles created and posted original content, other generic accounts played a supporting role by liking and sharing those posts. The exact reach and effectiveness of the campaign remain unclear, as officials did not disclose information about how many real users were exposed to the posts or whether they spread the messages further. The system evaded detection by X's authentication measures by automatically duplicating one-time passcodes sent to registered email addresses. Program code references to Facebook and Instagram hinted at the operation's plans to expand to those platforms. The intelligence agencies recommended that social media companies enhance their methods for identifying covert automated activities. X complied with a court order to provide information about the accounts to the FBI, subsequently deleting them. The company did not respond to inquiries from The Washington Post. The Justice Department expressed gratitude to X for its cooperation throughout the investigation, showcasing improved communication between the government and major social media companies following the Supreme Court's ruling affirming officials' right to expose foreign influence operations. John Scott-Railton, a researcher at the Canadian nonprofit organization Citizen Lab, stated that the detailed information shared by the countries about the botnet's operations aims to assist other investigators and companies in detecting similar activities. According to Scott-Railton, the documents reveal that Russian propagandists have utilized AI's large language models to expand their operations, aid in translation, and evade detection software designed to identify repeated use of IP addresses and other identifying markers.