This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A deepfake video impersonating the former Mongolian Deputy Prime Minister Amarsaikhan Sainbuyan was circulated on Facebook to promote a fraudulent investment scheme. The video, which used manipulated audio and lip movements, directed users to a website designed to harvest personal information. Fact-checkers confirmed the manipulation using Hive Moderation tools and identified the original footage as a 2024 interview.
A Facebook post reportedly used a deepfake-altered clip of then-Mongolian Deputy Prime Minister Amarsaikhan Sainbuyan to promote investment in the "Power of Siberia-2" pipeline and direct viewers to a registration website. The video allegedly modified a 2024 MONGOL TV interview with altered lip movement and added Russian audio, while the linked site sought personal information.
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.
The deepfake video was created to impersonate a trusted public official to deceive individuals into registering for a fraudulent investment scheme.
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 malicious actors who intentionally deployed deepfake technology to create a deceptive video for a financial scam.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
An AI-generated deepfake video of the Mongolian Deputy Prime Minister was used on Facebook to promote a fraudulent investment pipeline scheme and harvest personal data. The incident represents a minor national security concern, demonstrating how consumer-grade deepfake tools can be used to impersonate state officials for cybercrime and financial fraud.
The AI system used was a deepfake generation tool utilized by human creators to manipulate the video and audio of a politician. It operated as a tool, performing the specific function of altering lip movements and generating a voiceover based on direct human inputs.
Risk Level 3: Limited Risk. The incident involves AI-generated content, specifically a deepfake video, which falls under Limited Risk and is subject to transparency obligations under the EU AI Act.
The AI system was used to alter the lip movements of the politician in the video (Deepfake Video Generation) and overlay a synthetic Russian voiceover (Voice Generation).
The report from the Mongolian Fact Checking Center is clear, provides the source of the original video, and uses a verified tool (Hive Moderation) to confirm the deepfake with 99.9% confidence. The details of the scam and the data harvested are explicitly described.
None. The evidence clearly points to a deepfake-enabled phishing and financial scam.
Technology
Deepfake Generation Tools
AI tools allowed realistic alteration of lip movements and voice dubbing.
Lack of Real-time AI Filters
Social media platforms failed to block the deepfake at the upload stage.
Data Inputs
Unsecured Public Video Data
An official TV interview was harvested to train and generate the deepfake.
Human Factors
Exploitation of Public Trust
Using a high-profile official made the fraudulent investment scheme credible.
Limited User Awareness
Viewers struggled to identify synthetic media and fell for the scam.
Process and Methods
Delayed Moderation Action
Fact-checking occurred only after the post gained hundreds of reactions.
Regulatory Environment
Inadequate Deepfake Controls
Lack of strict regulatory standards for labeling synthetic media online.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
