This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Optometrist Joseph Allen reported that his likeness and voice were used in AI-generated deepfake videos on TikTok to fraudulently promote unverified myopia-reversal eyedrops. The scammers impersonated him to spread health misinformation and discourage professional eye care, with the fraudulent content persisting despite takedown efforts. The incident highlights the use of AI-generated content for identity theft and large-scale health fraud.
Optometrist and online educator Joseph Allen ("Dr. Eye Health") reported that AI-generated videos using his likeness and voice were circulated on TikTok to promote eyedrops purportedly claiming to reverse myopia. The alleged scam misrepresented Allen's views and discouraged visits to eye-care professionals. They reportedly resurfaced after a takedown.
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 AI-generated content promoted potentially dangerous, unverified medical products to consumers, posing a health risk.
The deepfakes generated and spread false health information, specifically promoting unverified eyedrops and falsely claiming the optometry profession is a scam.
According to boundary rule B1, because the scammers exploited Dr. Allen's public likeness and voice to deceive consumers for financial gain through product sales, this is classified under 4.3 rather than 2.1.
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 risk and resulting harm were caused by malicious human actors intentionally deploying AI deepfake generation tools post-deployment to commit fraud.
Added categories peak at 3: Substantial (Epistemic Harm, Inferred)
An optometrist's likeness and voice were stolen using AI deepfakes to promote unverified eye drops. While this represents a serious case of commercial fraud, identity theft, and health misinformation, it has negligible to minor national security implications and is manageable under standard law enforcement and platform moderation procedures.
The AI systems used in this incident (deepfake video and voice generators) operated as tools, performing specific, well-defined functions of generating synthetic media based on direct inputs provided by human scammers.
Limited Risk: The incident involves AI-generated content such as deepfakes, which are subject to specific transparency and disclosure obligations under the EU AI Act.
The primary technology used by the scammers was deepfake video and voice generation to synthesize Dr. Allen's face and speech for the promotional videos.
The reports provide clear, firsthand accounts from the victim, Dr. Joseph Allen, detailing the nature of the deepfakes, the platforms involved, the specific products promoted, and the operational challenges in getting the content removed. The facts align directly with the boundary rules for malicious scams using public likenesses.
None. The incident is a clear-cut case of commercial deepfake fraud and identity impersonation.
Technology
Advanced Deepfake Generation
AI tools generated realistic videos of the doctor's face and voice.
Algorithm-Driven Distribution
Social media engagement algorithms amplified low-effort AI slop videos.
Flawed AI Video Detection
Platforms lacked automated detection to flag synthetic medical content.
Data Inputs
Unprotected Public Video Data
Scammers used public YouTube videos to train voice and face models.
Human Factors
Limited Consumer Media Literacy
Viewers struggled to identify subtle deepfake visual and audio anomalies.
Trust in Authority Figures
Patients trusted the fake doctor's advice over professional care.
Process and Methods
Slow Platform Takedown Process
It took weeks for TikTok to verify identity and remove the scam videos.
Ineffective Deplatforming Rules
Scammers bypassed bans by creating new accounts in hours.
Management
Inadequate Platform Moderation Policy
Social media platforms failed to prioritize medical deepfake prevention.
Regulatory Environment
Lack of AI Identity Protections
No swift legal mechanism to stop unauthorized AI likeness cloning.
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.
