This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A 22-year-old B.Tech graduate was arrested in New Delhi for using AI image-generation tools to create and circulate obscene morphed images of a female college acquaintance to harass and humiliate her.
Delhi police alleged that a man named Keshav Bansal used photographs from a college acquaintance's private social media account with AI image-generation and editing tools to create obscene morphed images and upload them to anonymous Discord servers. The woman reported fake accounts and threatening messages intended to humiliate and harass her. Police arrested Bansal and said the investigation was ongoing.
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 public/semi-public circulation of obscene and explicit synthetic content on Discord and Instagram.
The perpetrator used AI tools to create humiliating and obscene imagery of a specific victim to target, harass, and manipulate her reputation, which fits sub-domain 4.3 under boundary rule B1.
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 a human actor who intentionally used deployed AI tools to generate and circulate non-consensual explicit imagery of a victim.
Added categories peak at 2: Minor (Psychological Harm, Direct)
A localized cyber-harassment incident in India involving the creation of non-consensual explicit deepfakes of a single victim. The perpetrator was arrested by local cyber police. The incident has negligible national security implications.
The AI system was used as a tool to perform a specific, well-defined function (image generation and editing) based on direct inputs (the victim's photos) without autonomous planning or context retention.
Risk Level 3. Limited Risk: The report describes the creation of AI-generated content such as deepfakes, which are subject to specific transparency obligations under the Act.
The perpetrator used AI-based image-generation and editing tools to create obscene morphed images of the victim using her facial images.
The reports provide clear, consistent, and detailed factual accounts of the incident, including the arrest of the perpetrator, his motives, and the specific actions taken using AI tools. There is no conflicting information between the two sources.
None
Technology
AI image-generation tools
AI-based editing tools allowed creation of obscene morphed images.
Anonymity of online platforms
Fake profiles and Discord servers allowed the creator to hide identity.
Data Inputs
Private social media photos
The perpetrator obtained the victim's photos from her social media account.
Victim's facial images
Facial images were used as inputs for generating obscene content.
Human Factors
Personal resentment
Perpetrator harbored a college grudge and sought revenge for insults.
Intent to harass
The perpetrator aimed to humiliate and threaten the victim online.
Malicious online community
Other users created and shared even more explicit morphed images.
Process and Methods
Multi-account creation
Operating multiple fake accounts and Gmails to distribute content.
Reposting explicit content
The perpetrator reshared additional explicit content made by others.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
