This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Public figures in Montenegro and Bosnia and Herzegovina have been impersonated in AI-generated deepfake videos and images to promote fraudulent health products and investment schemes. These scams, facilitated by platforms like Facebook and Instagram, have led to significant financial losses for citizens and have proven difficult for authorities to investigate due to the use of international servers and sophisticated digital concealment techniques. Experts highlight the urgent need for better platform detection and international cooperation to mitigate the rising tide of AI-enabled digital fraud.
Purported deepfake videos and images impersonating Montenegrin MP and surgeon Vladimir Dobričanin and other public figures were reportedly used in Facebook and Instagram ads to promote fraudulent health products and investment schemes in Montenegro and Bosnia and Herzegovina. The impersonations allegedly directed users to scam websites linked to firms such as Digital Edge and Nearest Edge, reportedly causing harm while evading timely platform enforcement.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to gain a personal advantage over others such as through cheating, fraud, scams, blackmail or targeted manipulation of beliefs or behavior. Examples include AI-facilitated plagiarism for research or education, impersonating a trusted or fake individual for illegitimate financial benefit, or creating humiliating or sexual imagery.
Also classified as
The deepfakes generated false statements and fake endorsements, leading users to believe public figures recommended these products.
The incident involves scammers using AI-generated deepfakes of public figures to execute financial scams, fraudulent health product sales, and fake investment schemes.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by human scammers intentionally using deployed AI deepfake generation tools to commit financial fraud and identity theft.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
Scammers utilized AI-generated deepfakes of public figures to conduct financial and health-related fraud in Montenegro and Bosnia and Herzegovina. While causing notable local financial losses and eroding public trust, the incident remains a localized cybercrime issue with minor national security implications.
The report describes the use of 'AI-based tools' to manipulate images, voices, photos, or videos of public figures. These deepfake generation systems operate on direct human input to produce synthetic media, functioning as tools for the scammers.
Risk Level 3. Limited Risk: The report describes 'AI-generated audiovisual content abuse' and 'deepfake techniques'. The EU Artificial Intelligence Act requires clear labelling of AI-generated synthetic content, which applies to deepfakes (Limited Risk).
The primary technology used in the scam was deepfake video and audio generation to impersonate public figures and simulate endorsements.
The report provides clear, detailed accounts of the deepfake scams, including specific names of affected public figures, companies involved (Digital Edge, Nearest Edge), financial loss statistics from the police (300,000 euros), and statements from official bodies like CIRT. The role of AI as a tool for generating deceptive synthetic media is explicitly established.
None. The incident is clearly a case of AI-enabled financial fraud and identity impersonation.
While the specific deepfake generation software names are not provided, this detail is not necessary for high-confidence classification under the taxonomies.
Technology
Widespread AI Tool Availability
Easy-to-use AI tools since 2022 enable rapid creation of deepfakes.
Convincing Deepfake Realism
Advanced AI generates highly realistic and persuasive simulated media.
Data Inputs
Misuse of Public Digital Identities
Unauthorized use of public figures' photos, videos, and voices as bait.
Human Factors
User Susceptibility to Deepfakes
Average social media users find fake videos highly real and convincing.
Urgency-Driven Decision Making
Users are lured by promises of quick wealth and act impulsively.
Process and Methods
Slow Platform Moderation Response
Delays in removing reported fake content allow scams to spread widely.
Use of Evasive Technical Methods
Scammers use VPNs, encryption, and foreign servers to hide traces.
Lack of Detection Tools
Sixty percent of organizations lack adequate deepfake detection mechanisms.
Management
Inadequate Platform Accountability
Social media platforms refuse to assess legality of reported fake profiles.
Absence of Pre-Publication Checks
Platforms fail to detect and flag deepfakes at the moment of upload.
Regulatory Environment
Weak International Cooperation
Cross-border server locations hinder swift investigation and prosecution.
Complex Legal Investigations
Technical and legal complexities make identifying perpetrators extremely slow.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
