In a surprisingly detailed composition Advice, agencies in the United States, the Netherlands, and Canada identified various software programs used to manage the network, including one named Melioter, which created fictitious users known as “ghosts” in various countries. The FBI won in court order The function allows us to capture the two web domains used to register the email addresses behind the accounts.
“Today’s actions are the first to disrupt a Russian-backed AI-enhanced social media bot farm,” said FBI Director Christopher A. Wray said. Report. “Russia intends to use this bot farm to spread AI-generated foreign disinformation to undermine our partners in Ukraine, influence geopolitical narratives favorable to the Russian government, and scale their mission with the help of AI.”
Automated accounts with more detailed biographies posted original content, while supporting actors from more generic accounts liked and reshared those posts. Officials did not respond to questions about how many actual users saw the posts and whether anyone spread the word further, so it’s unclear how effective the campaign was.
get caught
Stories that inform you
The system bypasses one of X’s techniques for verifying users’ authenticity by automatically copying one-time passwords sent to registered email addresses. References to Facebook and Instagram in the program’s code are intended to expand to those platforms, officials said.
The agency recommended that social media companies improve their methods to catch covert automated behavior.
X complied with the court order provided information about the accounts to the FBI, which later deleted them. The company did not respond to questions from The Washington Post.
The Justice Department thanked X for cooperating during the investigation, a sign of better communication between the government and major social media companies after the Supreme Court upheld the right of officials to point to foreign influence activities.
John Scott-Railton, a researcher at the Canadian non-profit Citizen Lab, said those countries have provided detailed information about the botnet’s inner workings to help other investigators and companies know what to look for.
“They don’t think this problem is going anywhere, so they’re sharing it widely,” Scott-Railton said.
The documents show that the AI’s large language models have helped Russian propagandists measure their activity and aid in translation, he said. It also helps them avoid detection software that looks for repeated use of the same Internet protocol addresses and other identifiers.
But many other systems are already in place, and depending on what they detect and what they get, they’ll get better, Scott-Railton said. “It’s not even the tip of the iceberg,” he said. “It’s the tip of the iceberg.”
