Russia uses AI-generated deepfake videos to undermine Ukrainian morale
Russia is using AI-generated deepfake videos depicting Ukrainian soldiers complaining about poor conditions and criticizing military leadership, distributed widely across social media to undermine Ukrainian and international support for the war. The videos are increasingly difficult to detect and are reportedly deployed as part of a coordinated information warfare campaign.
What happened
Russia is using AI-generated deepfake videos to conduct information warfare against Ukraine. According to Sensity AI, an Amsterdam-based company specializing in AI image security, these fabricated videos have become increasingly realistic and are distributed across social media platforms.
The fake videos typically depict Ukrainian soldiers complaining about poor front-line conditions, inability to contact command systems, and criticizing military leadership. The videos follow a fixed production pattern and are designed to appear authentic while reducing visual artifacts.
Reported impact
Sensity AI reports the videos are designed to target three audiences: Ukrainian soldiers outside the front lines who may believe nearby units are collapsing, Ukrainian civilians who may lose support for the war after seeing soldier complaints about government and military leadership, and international audiences whom the videos aim to discourage from supporting Ukraine through depictions of corruption and hopelessness.
The fake videos have established a consistent distribution pattern, typically starting on TikTok before being rapidly reposted to Telegram channels used in Russia and Ukraine, then amplified by media outlets, influencers, and military bloggers before spreading to Facebook, X, Instagram, and YouTube.
Index analysis
Sensity AI identified approximately seven types of AI propaganda videos, with fake impersonations of Ukrainian soldiers being the most common. The high similarity in filming style, language, and scenarios across multiple videos suggests a coordinated operation by Russian state actors.
Deepfake detection is becoming increasingly difficult as AI-generated videos avoid traditional artifacts like unnatural eye movement. The simple production setup—a single soldier speaking directly to camera—is deliberately designed to minimize visual flaws and make videos harder to identify as fake.
Governance controls
TikTok's relatively loose content moderation has made it the initial distribution platform, as videos can accumulate significant reach quickly. Once videos are widely reposted, platform content removal efforts cannot prevent further dissemination.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.