This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A series of deepfake advertisements featuring prominent Bulgarian public figures, including actor Kamen Donev, have been circulating on Facebook. These AI-manipulated videos use synthetic audio and altered lip movements to falsely endorse joint-pain products. The ads redirect users to fraudulent websites imitating legitimate news outlets to facilitate the sale of unverified medical products.
A reportedly AI-manipulated Facebook advertisement appeared to repurpose footage of Bulgarian actor, director, and playwright Kamen Donev, allegedly overlaying fabricated audio and altered lip movements to market a purported joint-pain product. The ad reportedly linked to a presumably fabricated article imitating a Bulgarian news outlet and to an online sales funnel for a product described as unavailable through pharmacies.
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 campaign generated and spread false medical claims and fabricated news articles to mislead consumers.
The incident involves the use of AI-generated deepfakes to impersonate public figures and falsely endorse products for financial gain, which constitutes fraud and targeted manipulation.
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 malicious human actors who intentionally deployed AI voice and video manipulation tools post-deployment to execute a financial scam.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
A localized deepfake scam campaign in Bulgaria used AI-generated video and cloned voices of public figures to sell unverified medical products. While demonstrating the ease of synthetic media creation, the incident is a commercial fraud scheme with negligible national security implications.
The AI system was used as a tool to generate synthetic voice and alter lip movements in existing video footage. The report states that the audio was manipulated with a 98% probability of being AI-generated, indicating the technology was used to perform a specific, user-directed function of voice replication.
Risk Level 3. Limited Risk: The report describes the creation and distribution of deepfakes, which are subject to transparency obligations under the EU AI Act to ensure users are aware they are interacting with AI-generated content.
The AI system was used to clone the voices of Bulgarian celebrities and synchronize their lip movements to falsely endorse a product.
The fact-checking report from Factcheck.bg provides clear, verified details about the deepfake nature of the videos, the specific celebrities targeted, the platform used for distribution (Facebook), and the verification tool used (ElevenLabs Audio Detection). The mechanics of the scam are well-documented.
None
Technology
AI Deepfake Voice Generation
ElevenLabs or similar tools enabled high-accuracy voice cloning of actors.
Video Lip-Sync Manipulation
AI manipulated mouth movements to match the fabricated audio tracks.
Data Inputs
Unauthorized Media Scraping
TV appearances of celebrities were harvested to train and feed AI models.
Human Factors
Exploitation of Public Trust
Scammers leveraged the high public reputation of well-known Bulgarian figures.
Consumer Vulnerability
Targets with joint pain were easily misled by trusted celebrity endorsements.
Process and Methods
Inadequate Ad Moderation
Facebook's sponsored ad screening failed to detect deepfakes before publication.
Spoofed Media Websites
Creating fake clones of Fakti.bg to deceive users into buying products.
Management
Facebook Ad Monetization Priorities
Platform prioritized ad revenue over rigorous pre-screening of sponsored ads.
Inadequate Impersonation Oversight
Lack of proactive policies to protect public figures from identity theft.
Regulatory Environment
Lack of Deepfake Ad Controls
Insufficient regulatory enforcement on social networks regarding deepfake ads.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
