A Russian propaganda network is promoting an AI-manipulated Biden video


Prominent accounts sharing the video include Russian Market, which has 330,000 followers and is run by Swiss social media personality Vadim Loskutov, known for praising Russia and criticizing the West. The video was also shared by Tara Reade, who moved to Russia in 2023 to seek citizenship. Reade also accused Biden of sexually abusing her in 1993.

The videos were also manipulated to avoid online detection, the researchers told WIRED. “Doppelganger operators trimmed the videos at arbitrary points, so they technically differ by milliseconds and can therefore be treated as separate unique videos by abuse-protection systems,” researchers at AntiBot4Navalny told WIRED.

“It’s unique because of its ambiguity,” Fink said. “It might be a well-known Russian band, but might not, might be a deepfake, but might not, might contain references to other politicians, but might not. In other words, it’s a typical Soviet-style propaganda video. The ambiguity allows for multiple competing versions, which means hundreds of articles and arguments online, which ultimately leads to more people watching it.”

With the Kremlin’s efforts to undermine the US election in November, it’s becoming clear that Russia is willing to use emerging AI technologies. A new report published this week by threat intelligence company Recorded Future highlights this trend, revealing that a Kremlin-linked campaign is using generative AI tools to push pro-Trump content on a network of fake websites.

The report details how a campaign called CopyCop used AI tools to rip content from real news websites, repurpose the content with a right-wing bias, and republish the content on a network of fake websites with names like Red State Report and Patriotic Review that claim to employ over 1,000 journalists — all of which are fake and invented by AI.

Topics raised by the campaign included mistakes made by Biden during speeches, Biden’s age, poll results that showed Trump in the lead, and claims that Trump’s recent criminal conviction and trial were “ineffective” and “totally messed up”.

It’s still unclear how much impact these sites are having, and a review by WIRED of social media platforms found very few links to the network of fake websites created by CopyCop. But what the CopyCop campaign has proven is that AI can amplify the spread of misinformation. And experts say it’s likely just the first step in a broader strategy that will likely include networks like Doppelganger.

“Engagement with websites is a difficult task to predict,” Clement Briens, an analyst at Recorded Future, told WIRED. “The AI-generated content may not be garnering any attention at all. However, it serves the purpose of helping these websites establish themselves as trusted assets when they publish targeted content like deepfakes (which) is amplified by established Russian or pro-Russian influential actors with an existing following and audience.”


Please enter your comment!
Please enter your name here

Share post:




More like this

How IT departments coped with the CrowdStrike chaos

Just before 1:00 a.m. local time on Friday, a...

Alphabet will invest an additional $ 5 billion in Waymo

Alphabet will spend an additional $5 billion on its...