Pro-Russian accounts use AI-generated videos showing fake Ukrainian soldiers surrendering to spread disinformation amid ongoing battle for Pokrovsk.
WARSAW: As Ukraine battles to defend the key eastern city of Pokrovsk, pro-Russian social media accounts are waging an AI-powered disinformation war.
Viral AI-generated videos depict Ukrainian soldiers surrendering weapons or weeping while retreating, despite the ongoing conflict.
Russia has been attempting to capture the Donetsk region logistics hub for over a year, intensifying its assault in recent weeks.
Battlefield maps from the Institute for the Study of War show Russian forces closing in on Pokrovsk’s outskirts in a pincer movement.
The fake videos accumulated millions of views across social media platforms in November.
Analyst Pablo Maristany de las Casas said they fit “a broader narrative we’ve seen since the beginning of the invasion”.
He noted the false claim that “Zelensky sending the young and elderly to the front line against their will because they are not doing well”.
Researcher Carole Grimaud stated the videos “instrumentalise uncertainty to sow doubt in public opinion”.
Visual inconsistencies reveal the AI-generated content, including a soldier walking easily despite a leg cast and a levitating stretcher.
Other fake videos bearing OpenAI’s Sora logo show soldiers in Ukrainian uniforms begging not to be sent to the front.
Some videos used the faces of Russian online streamers without their consent.
Exiled Russian YouTuber Alexei Gubanov confirmed his likeness appeared in a fake weeping soldier video.
He lamented that “a lot of people believe this… and that plays into the hands of Russian propaganda”.
The European Digital Media Observatory has published over 2,000 Ukraine war fact-checking articles since 2022.
Russian propaganda specialist Ian Garner called disinformation “an old tactic, but the technology is new”.
He explained the videos work by “chipping away at Ukrainian morale” while boosting Russian morale.
TikTok deleted accounts behind the videos after they garnered over 300,000 likes and millions of views.
OpenAI conducted an investigation but the videos continue circulating on Instagram, Telegram, Facebook and X.
They appear in Greek, Romanian, Bulgarian, Czech, Polish and French across multiple platforms.
Grimaud noted that “when it is repeated, it is possible that people’s perceptions change”.
AI chatbots also promote pro-Kremlin talking points, with one-fifth citing Russian state sources.
Maristany de las Casas warned that “the scale and impact of information warfare outpace the companies’ responses”.
