This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Activist and anti-image-based abuse advocate [name withheld] was targeted by a deepfake attack where her likeness was superimposed onto pornographic content. This incident, which followed her advocacy work against unverified content on PornHub, resulted in significant online abuse, doxing, and threats of violence against her.
Kate Isaacs, a London-based activist and founder of the #NotYourPorn campaign, was targeted in a deepfake incident. Her face was alleged to have been digitally manipulated onto a pornographic video using AI and shared online. The reported video, tagged with her name, is alleged to have led to streams of abuse, doxing, and threats of violence. The attack reportedly followed her efforts to pressure PornHub to remove unverified content.
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 victim and social media users were exposed to non-consensual sexually explicit and abusive content generated by AI.
The incident involves the targeted creation of humiliating and sexual deepfake imagery of a specific victim to silence and harass her, which fits the boundary rule for targeted humiliation under 4.3.
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 human actors intentionally using deployed AI deepfake tools to target, humiliate, and harass the victims.
Added categories peak at 4: Severe (Epistemic Harm, Inferred)
This incident involves the targeted creation of non-consensual deepfake pornography against an activist and a political candidate using consumer-grade AI tools. While causing severe psychological harm, doxing, and localized democratic disruption to a political campaign, the national security threat remains minor and manageable within standard law enforcement frameworks.
The AI deepfake software performs a specific, well-defined function of face-swapping or undressing images based on direct user input, operating without context retention or autonomous planning.
Limited Risk: The report describes AI-generated content such as deepfakes, which fall under Risk Level 3 and are subject to specific transparency obligations under the EU AI Act.
The AI systems were used specifically to generate deepfake videos and images by superimposing the victims' faces onto explicit bodies.
The reports provide detailed, consistent first-hand accounts from multiple victims regarding the creation, distribution, and impact of the deepfakes. The role of AI deepfake technology is clearly identified, and the resulting harms are well-documented across multiple reputable news sources.
None. The events clearly represent targeted harassment using synthetic media.
Technology
Advanced AI Face-Swapping Tech
Algorithms reconstruct highly convincing faces from minimal source footage.
Accessible Deepfake Applications
Apps like FaceMagic allow anyone to make deepfakes with a single click.
Data Inputs
Unprotected Social Media Photos
Innocent online photos and videos are harvested as source material.
Human Factors
Perpetrator Moral Disconnection
Creators dismiss psychological harm, viewing deepfakes as harmless fantasy.
Misogynistic Intimidation
Perpetrators weaponize deepfakes to silence and humiliate women campaigners.
Process and Methods
Reactive Content Moderation
Platforms rely on reports to remove content after the harm has occurred.
Flawed App Store Safeguards
Apps are distributed to minors without robust filters for sexual content.
Management
Inadequate Safety Resources
Tech companies fail to invest in proactive detection of sexual abuse content.
Regulatory Environment
Outdated Cyber Laws
Existing laws fail to criminalize non-consensual deepfake creation directly.
Delayed Online Safety Bill
Legislative delays leave victims without protection as technology evolves.
Limited Police Cyber Powers
Police forces lack the technical resources and powers to trace perpetrators.
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.
