AI-Powered Russian Bot Network Shuts Down by U.S. Authorities

The U.S. Department of Justice has seized two internet domains and uncovered nearly 1,000 social media Russian bot accounts promoting pro-Kremlin disinformation. These accounts had been used in an elaborate operation in which artificial intelligence technology was harnessed to create fake social media profiles. These profiles were often designed to look like they belonged to real Americans, with messaging that, in turn, supported the interests of the Russian government.

This network, comprising 968 accounts on the social media platform X, was managed by a Russian state-owned media outlet RT employee in collaboration with an officer from FSB Russia.

The operation began in April 2022 with the creation of a bot farm to advance Russian interests by misinforming the general public. The accounts were created and registered using private email servers hosted on two targeted domains and subsequently removed from the network due to violations of the terms of service.

The campaign worked simultaneously in several countries: the U.S., Poland, Germany, the Netherlands, Spain, Ukraine, and Israel. It was based on an AI-powered software package named Meliorator. According to the findings, the software in its most recent state was prepared for mass development and administration of artificial social media accounts devoted to spreading disinformation amongst a wide audience. The accounts were being created with authenticity in mind, using pictures and biographical information of people made up with an open-source application called Faker.

These bot personas can be divided into three primary categories: promoting political ideologies favorable to the Russian government, sharing messages from other bots, and propagating disinformation from both bot and non-bot accounts. The software used sophisticated techniques to avoid detection, such as replicating one-time passcodes or even using proxy IP addresses through countries with assumed residence of the AI-generated personas.

RT denied any wrongdoing and, without specifically addressing the accusations, criticized the characterization of farming as a common activity in Russia. This is believed to be the first time that the U.S. has publicly accused another government of using AI as a tool of a disinformation campaign. Criminal charges have not been laid, but the investigation is ongoing. In other related developments, tech giants such as Google, Meta, and OpenAI have all issued warnings with regard to the Russian disinformation operations. One of the most active networks has been an influential network called Doppelganger that made extensive use of platforms to distribute pro-Russian propaganda. This network works from large data centers in Europe and is linked with various cybercriminal activities.