This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Cherelle Kozak, an aspiring artist, was targeted by a scammer who used AI-generated video and voice to impersonate a famous rapper. The scammer attempted to trick Kozak into paying for fake radio promotion services. Kozak recognized the scam and did not provide any money, later confirming the fraud through a public warning issued by the celebrity being impersonated.
An aspiring artist in Austin, Texas, Cherelle Kozak, was targeted by a scammer using AI-generated video and voice to impersonate rapper Fat Joe. The impersonator appeared on a call, encouraged her to upload music for supposed radio play, and then demanded payment. Kozak did not comply. The scam closely resembled one that Fat Joe publicly warned about on January 5, 2025.
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 incident involved a scammer using AI-generated deepfake video and voice to impersonate a celebrity in a targeted attempt to defraud an individual.
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 a human scammer intentionally using deployed AI voice and video generation tools to execute an impersonation scam.
Added categories peak at 2: Minor (Epistemic Harm, Inferred)
An individual in Austin, Texas was targeted by an impersonation scammer using real-time AI-generated video and voice to mimic a famous rapper. The target identified the fraud, resulting in no financial loss. This incident represents a common consumer fraud attempt with negligible implications for national security.
The AI system was used as a tool to perform a specific function, namely mapping a celebrity's face and voice onto the scammer's input in real-time.
Risk Level 3: Limited Risk. The report describes the use of AI-generated deepfakes ('mimic your voice and your lips and make it look like you are seemingly saying something'), which falls under transparency obligations for AI-generated content.
The AI was used to generate a realistic deepfake video and clone the voice of a celebrity to impersonate them during a call.
The report clearly details the specific scam attempt, the technology used (AI voice and video mapping), the target, and the outcome (no financial loss). The role of AI is explicitly stated and the facts are unambiguous.
None
Technology
Real-time AI Video Mapping
Software maps celebrity faces and voices onto scammers during live calls.
Advanced Voice Mimicry
AI mimics voice and lip movements to make fake calls look realistic.
Data Inputs
Publicly Available Media Clips
Scammers use online video and audio clips of targets to train AI.
Minimal Data Requirements
AI software requires only a three-second clip to clone a person's likeness.
Human Factors
Inherent Trust in Video Calls
People naturally trust the face and voice they see on a live call.
Career Ambition Vulnerability
Victim's desire for a big break made her susceptible to the scam.
Urge to Act Quickly
Excitement leads to acting immediately without double-checking.
Process and Methods
Lack of Double-Checking
Failure to verify the identity of the virtual speaker before proceeding.
Direct Money Requests
Scammers immediately request money after establishing fake trust.
Regulatory Environment
Inadequate Prosecution Data
Lack of coordinated reports makes prosecuting scammers difficult.
AI system involved
An AI system was involved, but the reports do not identify it.
Identified by automated extraction; can differ from the alleged developers above.
