This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A Vero Beach man was arrested for using AI to generate and distribute non-consensual sexually explicit images of a female victim using photos taken from social media.
Vero Beach, Florida, police allege that Peter Solomon Ruma used AI nudification tools to create sexually explicit altered images of a former neighbor without her consent and distributed them through spoofed phone numbers. Investigators found more than 140 altered nude images of the woman on his phone. Ruma was arrested on September 2, 2026, and charged with six altered-depiction offenses, sexual cyberharassment, and stalking.
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 incident involved the creation and distribution of non-consensual sexually explicit deepfakes, which constitutes toxic and abusive content.
The incident involves a malicious actor using AI to generate non-consensual sexually explicit imagery of a specific victim for targeted harassment and humiliation, which falls under 4.3 according to the boundary rules (B1) for targeted exploitation and synthetic content abuse.
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 was caused by a human actor intentionally using a post-deployment AI tool to generate and distribute non-consensual sexually explicit images of a victim.
Added categories peak at 2: Minor (Psychological Harm, Direct)
A Florida man was arrested for using generative AI to create and distribute non-consensual sexually explicit images of a victim. While a severe violation of individual privacy and civil rights, the incident is a localized criminal matter with minor national security implications.
The AI system was used as a tool by the human operator to perform a specific, well-defined function (generating altered images from input photos) on direct input without autonomous planning or agency.
Limited Risk: The report describes the use of AI-generated content such as deepfakes, which falls under Risk Level 3 due to transparency and disclosure concerns regarding synthetic media.
The AI system was used specifically to generate realistic but fabricated sexually explicit images of real people using their existing photos.
The report provides clear, factual details regarding the arrest, the specific charges, the timeline of the offenses, and the nature of the AI's involvement in generating non-consensual explicit imagery.
None. The incident is a clear-cut case of non-consensual deepfake creation and harassment.
Technology
Generative AI Image Tools
Advanced AI tools enabled high-fidelity, realistic manipulation of images.
Lack of Deepfake Safeguards
AI software lacked built-in guardrails to block sexually explicit outputs.
Data Inputs
Unprotected Social Media Photos
Publicly accessible personal photos were harvested from social media.
Human Factors
Malicious Intent by Actor
The perpetrator deliberately harvested photos to harass and stalk the victim.
Prior Criminal Behavior Pattern
The suspect had a documented history of sexual offenses and burglary.
Process and Methods
Facilitated Online Distribution
Digital communication channels enabled the rapid sharing of deepfakes.
Inefficient Image Takedown Systems
Historically slow or complex mechanisms to remove nonconsensual imagery.
Regulatory Environment
Emerging Legal Frameworks
Laws and resources like TakeItDown are still developing to combat deepfakes.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
