This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Foreign scammers, identified as 'Yahoo Boys' and groups in Morocco, are using AI-generated images of U.S. military personnel to create fake social media profiles. These profiles are used to target vulnerable individuals, particularly widows, in long-term romance scams. Victims are manipulated into sending large sums of money via untraceable methods, with reported losses reaching hundreds of thousands of dollars per individual.
Yahoo Boys (from Nigeria and Ghana) and scammers from Morocco are reportedly targeting U.S. widows and vulnerable individuals using AI-generated images and fake military profiles in "Artificial Patriot" scams. They have allegedly impersonated military officials such as General Matthew W. McFarlane to gain trust, sharing fabricated backstories and emotional appeals. Once trust is established, they request money through untraceable methods.
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
AI-generated synthetic images were used to create highly convincing fake profiles that misled victims into believing they were communicating with real military personnel.
Scammers are using AI-generated images to create fake military profiles to conduct targeted romance scams and defraud vulnerable individuals of their life savings.
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 harm is caused by human scammers who intentionally use post-deployment AI image generation tools to create fake profiles and defraud victims.
Added categories peak at 4: Severe (Epistemic Harm, Inferred)
Foreign scammers are utilizing AI-generated synthetic images of U.S. military personnel to conduct highly effective romance scams. While the financial and psychological toll on individual victims is severe, the national security impact remains minor, primarily representing an evolved capability of transnational financial crime rather than a direct threat to state security.
The AI systems involved are image generation models used as tools by scammers to produce specific synthetic images of military personnel based on direct prompts.
Limited Risk: The report describes the creation and dissemination of AI-generated content such as deepfakes, which requires transparency obligations so users are aware they are interacting with synthetic media.
The AI systems were used specifically to generate synthetic, highly realistic images of military personnel and babies to create fake profiles for scams.
The reports provide detailed, consistent accounts of how scammers use AI-generated images to facilitate romance scams, including specific victim testimonies, financial loss figures, and statements from impersonated military personnel. The role of AI in generating synthetic images to bolster fake profiles is clearly described.
The scams could be viewed as traditional cybercrime where AI is merely an incidental tool, rather than an AI-driven safety incident.
Technology
AI Image Generation Tools
Generates realistic but flawed images of soldiers used as engagement bait.
Algorithmic Recommendation Engine
Facebook's algorithm served more AI images, amplifying scam reach.
Inadequate Automated Detection
Meta's systems failed to automatically detect and block AI-generated bait.
Data Inputs
Scraped Real Military Photos
Scammers harvested real generals' photos from government and social sites.
Public Relationship Statuses
Widow status on profiles allowed scammers to target vulnerable users.
Pre-written Scam Playbooks
Scripts and formats provided templates for convincing interactions.
Human Factors
Emotional Vulnerability
Widows and lonely individuals seeking companionship fell for romance scams.
Lack of Digital Literacy
Victims could not identify subtle AI image artifacts like extra fingers.
High Trust in Military Personnel
Patriotic victims naturally trusted profiles claiming to be U.S. soldiers.
Process and Methods
Inadequate Profile Verification
Platforms allowed easy creation of fake profiles posing as generals.
Flawed Impersonation Reporting
Real generals' authentic accounts were shut down instead of fake clones.
Whack-a-Mole Mitigation
Reactive removal of fake accounts was too slow to counter rapid creation.
Management
Prioritizing Engagement Over Safety
Meta's algorithm promoted engagement bait to increase user platform time.
Under-resourced Moderation Teams
Meta failed to allocate enough human reviewers to handle the scale of clones.
Defensive Corporate Stance
Meta suggested victims fix issues themselves instead of taking responsibility.
Regulatory Environment
Inadequate Law Enforcement Action
Local police and FBI struggle to investigate overseas cyber criminals.
Lack of Platform Accountability
No strict regulations holding platforms liable for hosting scam networks.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
