A newly identified AI-generated video targeting residents of Sumy has become the latest evidence of Russia’s expanding information offensive against Ukrainian border communities. The video, detected by Ukraine’s Center for Countering Disinformation under the National Security and Defense Council, shows an unidentified man who appears to be a local resident. In an emotionally charged appeal, the man claims that his apartment has no electricity or water supply, and he allegedly provides his first and last name to make the complaint seem genuine. According to the center’s official statement, the footage is not authentic but was generated using artificial intelligence. The disclosure was published on the center’s Telegram channel, which has become a key resource for exposing Kremlin disinformation. This particular forgery, however, is part of a broader pattern: rather than a random act by an individual, it is a coordinated psychological operation designed to destabilize a region that has already faced years of military pressure. The fact that the video was circulated on social media at a moment when residents were still coping with difficult conditions from the war is no coincidence. Ukrainian officials have repeatedly warned that Russian information troops treat the civilian social media space as a battleground, using fabricated stories about failed infrastructure, urgent crises, and alleged official negligence. The Center for Countering Disinformation emphasized that the man in the video is likely completely invented, down to his name and false appearance, and that the entire scenario was manufactured to convince Sumy residents that their utilities had collapsed and that no help was available. This early assessment is grounded in the recognition that Russian propagandists have learned to disguise their fake narratives in the familiar language of everyday complaints, hoping to bypass critical thinking by presenting an emotional, plausible-sounding grievance from a supposedly ordinary neighbor. The Sumy region, located in northeastern Ukraine, has been especially vulnerable to Russian attacks since the beginning of the full-scale invasion, and information warfare has long accompanied kinetic strikes there.
The fake video follows a well-known but still effective formula: a man calls himself an ordinary local and says he is forced to speak out because no one else will tell the truth. He appears to be standing in a dimly lit apartment, and his manner suggests exhaustion and frustration. He insists that there is no electricity and no water in his home, and he gives a full name, apparently calculating that a common surname will be difficult for audiences to verify in real time. The Center for Countering Disinformation quoted this detail specifically to explain how the deception was built. By attaching a name to the complaint, the creators of the fake hoped to make it look like an authentic grassroots video, the kind that might be posted by someone who is not a politician, not a journalist, and not a soldier, but simply a resident who is at the end of his patience. The emotional tone is deliberately chosen: anger and helplessness are more likely to provoke rapid sharing among viewers, especially in a community that has experienced real utility cuts and security threats. At the same time, the content of the complaint is not wholly impossible, which is what makes the forgery dangerous. Sumy and the surrounding region have suffered repeated attacks on energy infrastructure during the conflict, and power outages have occurred from time to time. A deepfake that exploits those real conditions can blur the boundary between fact and fiction, making it harder for ordinary residents to know what to believe. The Center’s message was intended to stop the video from spreading further and to reassure residents that regional authorities have not left them without water or electricity, and that the video is a manipulated production, not a report from a real neighbor. The near-realistic setting, the informal Russian language that the speaker uses, and the absence of any official insignia or verifiable dates all contribute to a dangerous illusion of authenticity. This is exactly the kind of targeted influence operation that Ukrainian analysts have learned to anticipate: it takes a genuine grievance from the population, adapts it to a fake story, and then injects the story into the widest possible media stream.
Experts at the Center for Countering Disinformation did not rely on intuition when classifying the video as a deepfake. According to the center’s statement, analysts used specialized services specifically designed to detect AI-generated content, and the review confirmed that the recording was constructed artificially. The telltale signs included delayed facial expressions, unnatural blinking, and a synthesized voice. Together, these anomalies indicate that the video was created using generative algorithms and not recorded by an actual person in an apartment. Facial movements were not synchronized with the emotional tone of the speech; the blink rate did not match the cadence of the speaker’s words; and the prosody of the voice, its intonation and stress patterns, had the flattened, mechanical quality often associated with text-to-speech models. Such errors are typical of currently available video-generation tools, which have become significantly more sophisticated but still leave subtle residues of their synthetic origin. The center’s warning is especially relevant because social media users often cannot distinguish a high-quality deepfake from a genuine video, particularly when the clip is short and recorded on a smartphone. In this case, the visual imperfections might be dismissed as the result of a poor internet connection or a cheap camera, but machine-analysis programs can spot inconsistencies that are invisible to the naked eye. The decision to make the apartment appear dim and the image slightly grainy was likely a deliberate effort to hide the generative flaws. The creators also made the video short, perhaps no more than 30 seconds, to minimize the opportunities for close inspection. The finding underscores a larger technological trend: the weapons of modern disinformation are no longer just anonymous text posts and written rumors; they are now lifelike synthetic media designed to impersonate citizens and officials alike. The Center for Countering Disinformation is used to addressing this evolving threat, and in recent years it has detected many similar deepfakes involving politicians, soldiers, and civilians. Each case helps Ukrainian institutions refine their methods of verification. The Sumy video fits into a growing library of known fakes, and it is possible that the same AI model used to generate this complaint has been used in other recordings across the country. The key lesson, experts say, is that emotional appeals perceived to be from ordinary people cannot be taken at face value simply because they look like home videos.
The distribution strategy observed in the Sumy case follows a familiar Russian operation pattern. The Center for Countering Disinformation reported that the video was initially uploaded from an anonymous TikTok profile. It then gained traction and was rapidly disseminated by several Russian propaganda Telegram channels, many of which have large audiences and coordinated posting schedules. This pattern is typical of hostile bot networks: a fake is launched in an anonymous space that is difficult to attribute, then the network of controlled channels amplifies it within a matter of hours. The goal is to create an impression of viral organic circulation, as if thousands of people independently began sharing the video because they recognized a real problem. In reality, the propagation is itself part of the fabrication. TikTok was chosen for the first release because its algorithm tends to promote short, emotionally charged videos to viewers based on engagement, not source credibility. The anonymous nature of the platform makes it easy to create a new account, upload a clip, and discard the account before it can be subjected to scrutiny. The Telegram channels then step in to give the video a political framing, attaching captions that blame Ukrainian authorities, regional administrations, and emergency services for the alleged catastrophe. The creators likely expected that after the video appeared in Russian-language channels, mainstream social media users and perhaps even some Ukrainian media outlets would pick it up as a real complaint. This amplification chain has been documented by analysts monitoring Russian disinformation in Ukraine for many years. It uses the anonymity of one platform as a shield and the relative openness of another as a megaphone. The result is a hybrid narrative that cannot be easily traced to a single source, because by the time journalists or official institutions respond to it, the video is already visible in dozens of distinct channels. In this particular case, the Center exposed the mechanism before the story became widespread. Its prompt intervention prevented the fake from being accepted as ground truth and allowed utility providers and local government agencies to issue denials while the evidence was still fresh. But the same distribution recipe is constantly reused, and there is no sign that Russia has stopped hiring propagandists, running bot farms, or paying algorithmic promotion services to spread these videos.
Analysts said the deeper purpose of the deepfake is not merely to attract attention but to produce measurable effects on civilian behavior and morale. The Center for Countering Disinformation warned that the operation was intended to exert psychological pressure on Sumy residents, to propagate a narrative of an inevitable humanitarian catastrophe, and to undermine trust in local authorities and emergency services. These goals are mutually reinforcing. If residents believe that the government has abandoned them and is hiding the truth about utility outages, they may lose confidence in public institutions and respond to the crisis by hoarding supplies, fleeing their homes, or ignoring official instructions. That kind of social disorder can cause damage that physical weapons cannot achieve alone. The video also creates the impression that the situation in Sumy is much worse than any official account, which supports the Russian information campaign aimed at foreign audiences, who may be convinced that Ukraine cannot manage its own territory. At the same time, the narrative is designed to demoralize residents by telling them they are forgotten and that their suffering is invisible. For a border region that has already experienced missile strikes, drone attacks, shelling, power shortages, and mass displacement, a message of abandonment can be deeply unsettling. The Center’s public identification of the deepfake is a direct countermeasure, giving residents a concrete explanation for why they may see the video and why they should not share it. The statement also implicitly reminded residents that Russia has a long history of staging AI videos across the Sumy region and the broader Ukrainian eastern and northern borderlands. These videos are not isolated creations; they are part of a continuous stream of information sabotage. Some previous fakes have shown officials making admissions, soldiers making threats, or civilians describing conditions that do not exist. The Sumy video fits into a pattern of Russian attempts to exploit the real hardships of the war while denying the reality that those hardships are caused by Russian attacks. In this way, the deepfake acts as a two-part weapon: first, it makes false claims about the collapse of utilities; second, it distracts attention from the simple fact that any genuine utility damage in Sumy has generally resulted from Russian strikes. By blaming Ukrainian authorities, the propaganda attempts to reverse the roles of aggressor and victim and to weaken the political support of the regional government.
The Ukrainian side responded to the deepfake not only with technical analysis but also with a broad appeal to media literacy and public vigilance. The Center for Countering Disinformation called on citizens to check information from official sources, to trust announcements made by local authorities and utility companies, and to report suspicious videos to monitoring organizations. This response reflects a key lesson learned during years of hybrid warfare: the best defense against disinformation is a well-informed public that refuses to share unverified emotional content. Residents of Sumy are advised to look for official confirmation before accepting any claim about water cuts, electricity outages, or emergency measures. Official Ukrainian websites, municipal Telegram channels, national energy operators, and local administrations are the only trusted locations for such operational updates. In addition, the Center encourages citizens to be especially suspicious of anonymous videos that call for immediate outrage, especially if the speaker mentions a full name but provides no verifiable address, occupation, or other details. The public can also support countermeasure efforts by reporting deepfakes to the Center’s Telegram bot or to fact-checking organizations, allowing analysts to investigate the source and issue a warning as quickly as possible. These measures are not meant to silence legitimate complaints; there are real problems, and residents have a right to voice them. But the distinction between truth and manipulation matters. The creation of a fake resident who claims to be without electricity and water is an attack on everyone who genuinely experiences utility failures, because it undermines the credibility of all future complaints. When a true report of hardship appears, it may now be dismissed as another deepfake. That is precisely the strategic goal of the propaganda: to poison the information space so thoroughly that citizens no longer know whom to trust. The Center’s prompt work shows that detection systems are working and that the propaganda machine has not been allowed to spread its message unopposed. Yet the fight is ongoing. Russian information operations against Sumy and other Ukrainian communities will continue to evolve, as will the tools used to expose them. The best way to defeat the panic, distrust, and despair that these deepfakes are designed to create is to combine professional verification with the resilience of each individual resident. For now, the most important message from the Center for Countering Disinformation is clear: do not share the video, rely on official channels, and remain aware that the visible face of Russia’s war is not only a missile or a drone; it is often a fake neighbor speaking into a smartphone, asking you to believe a lie.
