This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A Utah resident was defrauded of over $400 after purchasing weight loss supplements advertised through online videos featuring an AI-generated deepfake of Oprah Winfrey. The product received was primarily turmeric, rather than the advertised ingredients. The incident highlights the growing prevalence of celebrity deepfakes used in online scams.
A Utah woman, Lisa Swearingen, reported paying more than $400 for weight loss supplements after seeing online ads featuring a purported Oprah Winfrey deepfake endorsement. When the product arrived, she said its primary ingredient was turmeric rather than the advertised formula.
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 deepfake harms the celebrity's reputation and marketability by undermining the trust in their real words and endorsements.
The incident involves bad actors using an AI-generated deepfake of a celebrity (Oprah Winfrey) to run a fraudulent weight-loss supplement scam, directly targeting consumers for financial gain.
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 bad actors intentionally using AI deepfake technology post-deployment to defraud consumers.
Added categories peak at 3: Substantial (Epistemic Harm, Inferred)
A consumer fraud incident involving an AI-generated deepfake of celebrity Oprah Winfrey to sell weight-loss supplements. While highlighting the growing ease of creating deceptive synthetic media, the incident is a localized commercial scam with negligible national security implications.
The AI system was used as a tool to generate deepfake videos of Oprah Winfrey based on existing footage. The report notes 'it's easy for a bad actor to make a video of that person' using relatively little source material, indicating the AI functioned as a deterministic tool for content generation.
Limited Risk: The report describes the use of AI-generated deepfakes. According to the definitions, 'AI-generated content such as deepfakes' falls under Risk Level 3 (Limited Risk) which requires transparency obligations to inform users.
The AI was used to generate deepfake videos of Oprah Winfrey to be posted on TikTok and Reels.
The reports provide clear details about the scam, the victim's financial loss, the use of an Oprah Winfrey deepfake, and Oprah's public response. The role of the AI (deepfake generation) is explicitly stated, and the harm is well-documented.
None
Technology
Realistic Deepfake Generation
AI tools generate highly realistic celebrity deepfakes that are hard to spot.
Low Resource Requirements
Deepfakes can be created using relatively little source material.
Data Inputs
Abundant Public Video Data
Vast online footage of celebrities makes training deepfake models easy.
Lack of Content Provenance Data
No digital labels or metadata exist to verify if video is AI-generated.
Human Factors
Trust in Celebrity Likeness
Consumers trust endorsements from recognizable public figures like Oprah.
Difficulty Spotting Fraud
Consumers struggle to identify highly sophisticated AI-generated deepfakes.
Process and Methods
Inadequate Ad Verification
Social media platforms lack effective processes to filter deepfake scam ads.
Deliberately Vague Websites
Seller websites use vague details to obscure true product ingredients.
Management
Lack of Seller Responsiveness
Seller ignored customer complaints and evaded accountability.
Regulatory Environment
Lack of Labeling Regulations
Absence of laws forcing platforms to label AI-generated content.
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.
