This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
In October 2023, a 15-year-old male student at Aledo High School in Texas used AI-powered software to create and distribute non-consensual nude images of 14-year-old Elliston Berry and eight other female classmates. The images were shared via Snapchat, causing significant emotional distress and anxiety for the victims. The incident prompted the victims and their families to advocate for federal legislation, specifically the 'TAKE IT DOWN Act,' to criminalize the creation and distribution of such deepfake content and mandate its removal by social media platforms.
In October 2023, an unnamed male student at Aledo High School outside of Fort Worth, Texas allegedly generated and distributed deepfake nude pictures of classmate Elliston Berry and six other female students in her friend group via social media. Berry's mother, Anna McAdams, spoke with Senators Ted Cruz and Amy Klobuchar, who consequently drafted the Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks (or TAKE IT DOWN) Act.
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 content depicting minors, exposing classmates and social media users to toxic material.
The classmate used AI software to create humiliating and non-consensual sexual imagery targeting specific victims for personal harassment and social degradation.
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 classmate who intentionally chose to use AI clothes-removal software to generate and distribute explicit images of his peers.
Added categories peak at 3: Substantial (Child Sexual Exploitation & Abuse (CSEA), Direct)
A high school student used consumer AI clothes-removal software to generate and distribute non-consensual explicit deepfakes of classmates. While causing severe psychological harm to the victims and highlighting regulatory gaps in protecting minors from AI-enabled exploitation, the incident is a localized criminal and civil issue with minor national security implications.
The AI clothes-removal software performed a specific, well-defined function of manipulating uploaded images based on direct user inputs without independent planning, context retention, or autonomous decision-making.
Limited Risk: The report describes the generation of deepfakes, which are classified as limited risk under the EU AI Act and are subject to specific transparency and disclosure obligations.
The primary function of the AI software was to generate hyper-realistic deepfake nude images by manipulating existing non-explicit photographs of the victims.
The incident is documented across multiple highly detailed and consistent news reports, including direct testimonies from the primary victim, her mother, and involved lawmakers. The facts regarding the AI's role, the distribution platform, and the resulting harms are clear and uncontradicted.
None. The reports consistently describe the event as a targeted cyberbullying and sexual exploitation incident facilitated by generative AI.
Technology
Accessible clothes-removal software
Easy online tools allowed quick generation of realistic nude deepfakes.
Hyper-realistic image generation
Advanced AI made the doctored images indistinguishable from real photos.
Proliferation of generative AI
Rapid technological advances made powerful tools widely available to minors.
Data Inputs
Insecure social media photos
Classmate downloaded innocent public or shared photos from Instagram.
Lack of image protection tech
No mechanisms to prevent scraping or downloading of personal images.
Access to private accounts
The perpetrator gained access by following a private Instagram account.
Human Factors
Malicious intent of classmate
A male classmate chose to exploit and harass peers using AI.
Lack of consent awareness
Minors lack understanding of digital consent and consequences of deepfakes.
Misplaced social trust
Victims trusted acquaintances with access to private social media accounts.
Process and Methods
Slow social media takedowns
Snapchat ignored family requests for nine months until a senator intervened.
Inadequate platform moderation
Social media lacked automated processes to flag and block deepfake NCII.
Weak school incident response
The high school lacked protocols to handle AI-enabled sexual exploitation.
Management
Outdated student codes of conduct
School board failed to update policies to address AI-driven misconduct.
Platform negligence of user safety
Snapchat failed to prioritize and resource victim reporting mechanisms.
Inadequate school support systems
School failed to provide mental health support or protect target students.
Regulatory Environment
Absence of federal deepfake laws
No federal legislation mandated rapid removal or criminalized AI deepfakes.
Inconsistent state regulations
State laws were patchy, with some covering deepfake videos but not photos.
Law enforcement confusion
Police lacked clear legal frameworks and training to prosecute AI deepfakes.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
