This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Two men in Silleda, Spain, were identified by the Guardia Civil for creating and distributing non-consensual AI-generated sexual images of 22 women. The perpetrators used photos scraped from the victims' social media profiles to generate the deepfake content. The investigation was initiated following multiple complaints filed by the victims starting in June 2026.
Spain's Guardia Civil identified two men in connection with purportedly AI-generated sexual images depicting 22 women from Silleda, Galicia. Investigators said photographs taken from the women's social media profiles were altered to portray them nude or in sexual scenes; a 35-year-old allegedly created the material and a 29-year-old allegedly distributed it online. Victims began filing complaints in June 2026.
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 explicit, non-consensual sexual content, exposing victims and others to toxic material.
The perpetrators used AI deepfake tools to target specific victims for personal harassment and humiliation, creating explicit sexual imagery of them without consent.
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 the intentional actions of human perpetrators who used deployed AI tools to create and distribute non-consensual explicit images.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
Two individuals in Spain used AI deepfake tools to create and distribute non-consensual explicit images of 22 women. While representing a severe violation of individual privacy and human rights, the incident remains a localized criminal matter managed by domestic law enforcement with negligible direct implications for national security.
The perpetrators used AI applications as tools to manipulate existing photos and generate deepfake images. The AI performed a specific, well-defined function based on direct user input without autonomous planning or agency.
Limited Risk: The AI Act classifies AI-generated content such as deepfakes under Limited Risk, which imposes specific transparency obligations to ensure users are aware the content is AI-generated.
The AI system was used specifically to generate realistic deepfake images of the victims naked, using their social media photos as inputs.
The reports from multiple Spanish news outlets consistently detail the incident, including the number of victims (22), the location (Silleda), the age group of the victims (20-30), the arrest/identification of the two suspects, and the technology used (deepfake AI apps scraping social media). There is minimal ambiguity regarding the core facts.
None. The facts clearly point to a targeted non-consensual deepfake harassment incident.
Technology
AI Deepfake Applications
AI tools allowed easy creation of realistic nude images from normal photos.
Lack of Platform Filters
Messaging platforms failed to automatically detect and block deepfake content.
Data Inputs
Public Social Media Photos
Victims' photos were harvested directly from public social media profiles.
Human Factors
Malicious Intent of Creators
Perpetrators deliberately generated and shared non-consensual sexual content.
Low Privacy Configuration
Victims did not restrict public access to their personal profile pictures.
Process and Methods
No Consent Verification
AI apps did not verify user consent before modifying uploaded facial images.
Rapid Digital Distribution
Messaging networks facilitated immediate sharing of the generated images.
Management
Weak App Safety Controls
AI developers failed to restrict their models from generating nude outputs.
Regulatory Environment
Weak Legal Safeguards
Insufficient preemptive regulations on the misuse of AI deepfake platforms.
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.
