Skip to Content
logologo
AI Incident Database
Open TwitterOpen RSS FeedOpen FacebookOpen LinkedInOpen GitHub
Open Menu
Discover
Submit
  • Welcome to the AIID
  • Discover Incidents
  • Spatial View
  • Table View
  • List view
  • Entities
  • Taxonomies
  • Submit Incident Reports
  • Submission Leaderboard
  • Blog
  • AI News Digest
  • Risk Checklists
  • Random Incident
  • Sign Up
Collapse
Discover
Submit
  • Welcome to the AIID
  • Discover Incidents
  • Spatial View
  • Table View
  • List view
  • Entities
  • Taxonomies
  • Submit Incident Reports
  • Submission Leaderboard
  • Blog
  • AI News Digest
  • Risk Checklists
  • Random Incident
  • Sign Up
Collapse

Report 5394

Associated Incidents

Incident 11192 Report
Alleged Deepfake Videos Impersonate Lithuanian Politicians and Doctors in Purported Cross-Border Scam Network

Loading...
Deepfakes target Lithuania as part of wider European scam campaign
lrt.lt · 2025

A series of highly convincing fake news videos circulating online this month startled residents in Lithuania. Created using artificial intelligence technology known as deepfakes, the fraudulent videos are believed to be part of an international scam network targeting multiple European countries.

The deepfakes mimicked familiar Lithuanian television news programmes, complete with fabricated interviews featuring journalists, politicians, and doctors. In one video, a deepfake version of MEP and former Health Minister Aurelijus Veryga appears to promote an eye health product.

"If in 2025 you are still putting the first eye drops you find into your eyes, you are either ignorant or behaving like complete idiots," says the Veryga deepfake.

Another video shows doctor Vytautas Kasiulevičius claiming vaccines harm health while promoting a medicinal product.

Operating across Europe

The manipulated videos were disseminated via Facebook, where they reached audiences in several EU countries. According to Viktoras Daukšas, head of the fact-checking initiative Debunk.org, researchers found 20 videos, which have now been blocked.

Experts noted that such high-quality fakes had not been seen in Lithuania before.

"The errors are visible only in some grammatical cases and stress patterns. This campaign uses deepfake videos of the highest quality we have detected in Lithuania so far," said Daukšas.

He explained the videos were created using publicly available footage of public figures and politicians from social networks.

The National Crisis Management Centre called the campaign a "new level of fraud" and expressed concern over the advanced nature of the fakes.

According to Daukšas, scammers relied on sensational headlines about vaccine harm to attract attention. Anti-vaccine activists then spread the scam links designed to steal money.

"This increased visibility -- more people were robbed," he noted.

Daukšas added that the fraudulent content had likely been seen by hundreds of thousands of people across the European Union.

"The numbers are large, these deceptive ads have been shown in about 14-15 EU countries. Over 20 accounts have already been found, and more are discovered daily. I think we're talking about hundreds of thousands [of views] or more," he said.

Scammers' tactics evolving

Tomas Stamulis, IT security lead at Surfshark, noted scammers had used similar tactics before -- posing as media outlets and creating fake articles. But now the fakes are very realistic and faster to produce. AI's understanding of Lithuanian is also improving.

"A lot of work was needed before from designers, tech experts, engineers to make the person in the report look real. Now it takes only a few AI tools, the right prompts, some photo samples, and you have a report that looks quite realistic and can fool people," he said.

Stamulis explained that traditional giveaways of fake AI-generated imagery -- such as distorted ears or fingers -- were no longer reliable. In many cases, the manipulated videos appear flawless to the casual viewer.

"It's very important to think critically -- ask yourself, do you see information from just one source or many portals and people? Where does the link lead? Is it really, for example, the official LRT site, or something similar but unrelated -- that could give away the source," he advised.

When watching videos, he suggested paying attention to video and audio dynamics -- abrupt cuts, pauses, or small glitches can be clues. But scammers only need one victim.

"Groups like this, when trying to trick ordinary people, don't spend much time on content, message, or the actual image, video, or audio. Because they know that out of a thousand people, a few dozen will fall for it anyway," Stamulis emphasised.

Lack of social media accountability

The experts also raised concerns about law enforcement's limited ability to tackle these crimes. Many of the perpetrators operate from countries such as Russia, which does not cooperate with Lithuanian authorities, or Ukraine, where ongoing war complicates investigation.

"Money from Lithuanian accounts is quickly transferred to foreign banks, and it's very difficult to stop or block such funds. Often people realise they have lost money later, thinking for a long time it was an investment or return -- but nothing comes back," Stamulis noted.

The fake videos have now been removed from social media. The National Crisis Management Centre's senior adviser Darius Buta said social media platforms' approach remains the biggest concern for state institutions.

"We tried contacting Facebook's representative for the Baltic states, they promised to strengthen content review, but so far there are no results," he told ELTA earlier.

Experts agree that addressing such scams requires coordinated cross-border action.

"Law enforcement's job is to continue analysis. We found this is an international group and operation; investigating these cases requires international efforts. Lithuania's police alone cannot solve them," said Daukšas.

Read the Source

Research

  • Defining an “AI Incident”
  • Defining an “AI Incident Response”
  • Database Roadmap
  • Related Work
  • Download Complete Database

Project and Community

  • About
  • Contact and Follow
  • Apps and Summaries
  • Editor’s Guide

Incidents

  • All Incidents in List Form
  • Flagged Incidents
  • Submission Queue
  • Classifications View
  • Taxonomies

2024 - AI Incident Database

  • Terms of use
  • Privacy Policy
  • Open twitterOpen githubOpen rssOpen facebookOpen linkedin
  • e1b50cd