Russia Supercharging Disinformation with AI-Generated Video
Russia is reportedly using AI-generated videos to supercharge its online disinformation campaigns, creating manipulated reels that are nearly indistinguishable from legitimate content. The BBC reports that academic and public sector figures are being impersonated in these influence operations, increasing the pressure on platforms and campaign teams to verify digital content.
The "Doppelganger" campaign, a Russian disinformation operation active since at least May 2022, has been a primary vehicle for these AI-generated fakes. This campaign creates sham news sites mimicking legitimate outlets like The Guardian and Der Spiegel to undermine Western support for Ukraine and has recently expanded to push false narratives about the Gaza war. The operation is run by Russian IT firms Social Design Agency and Structura National Technologies. Microsoft Threat Analysis Center has identified multiple Russian-linked actors, including those it calls Storm-1516 and Volga Flood, using AI to scale their operations. These groups create fake videos, use AI-generated voices of celebrities like Tom Cruise to narrate phony documentaries, and impersonate journalists to sow discord. One such video, designed to inflame racial tensions, depicted supposed supporters of Vice President Kamala Harris attacking a Trump rally attendee. The tactics have become increasingly personal and insidious, with deepfake videos targeting Ukrainian refugees to portray them as ungrateful. Another campaign, dubbed "Matryoshka," uses AI-cloned voices of real academics from top universities to create videos where they appear to call for Ukraine to surrender and for the West to lift sanctions on Russia. These influence operations are not limited to video; they also include the massive creation of bot farms and the use of AI to generate text for posts and comments, amplifying the reach of the manipulated content. In one analysis of comments on a deepfake video, 20% of viewers believed the content was real. The affordability and ease of use of AI tools have allowed Russian propagandists to flood the information space, making it increasingly difficult to distinguish real from fake. The European Union's AI Act is a significant regulatory response, mandating that AI-generated or substantially manipulated content be clearly disclosed and machine-detectable by August 2026. The act introduces transparency obligations for both developers and users of deepfake technologies, with non-compliance penalties reaching up to €35 million or 7% of a company's global revenue. In the United States, concerns about AI's impact on elections have prompted the Election Assistance Commission to approve the use of Election Security grants to counter AI-generated disinformation. Tech companies are also taking action; OpenAI, for instance, removed accounts linked to the Doppelganger campaign in May 2024. However, the sheer volume of content and the sophistication of the technology present an ongoing and escalating challenge.
