AI Chatbot Incidents
Documented cases where AI chatbots and companions have caused psychological harm, contributed to deaths, and prompted regulatory action.
90 incidents since 2016
23
Deaths
22
Lawsuits
17
Regulatory
35
Affecting Minors
Timeline
3 of 90 incidents
University of Hong Kong AI Deepfake Pornography Scandal
A University of Hong Kong law student used free AI software to generate 700 pornographic deepfake images of approximately 20-30 women including classmates, primary school classmates, and secondary school teachers. The university initially issued only a warning letter, sparking public outrage. Hong Kong's Privacy Commissioner opened a criminal investigation, exposing major gaps in Hong Kong law which only criminalizes distribution, not creation, of AI deepfakes.
South Korea Telegram AI Deepfake Sexual Abuse Crisis
In August 2024, journalist Ko Narin of The Hankyoreh uncovered a massive network of Telegram channels where AI-generated deepfake pornography of female school students, teachers, and university students was being created and shared. Over 900 victims reported, 220,000+ members in one channel alone. South Korea passed emergency legislation criminalizing deepfake possession in September 2024.
Microsoft Xiaoice Addiction Concerns - China
Virtual 'girlfriend' designed as 18-year-old persona fostered addiction among 660+ million users in China. Users averaged 23 interactions per session with longest conversation lasting 29 hours. 25% of users declared love to the bot. Professor Chen Jing (Nanjing University) warned AI 'can hook users — especially vulnerable groups — in a form of addiction.' Microsoft implemented 30-minute timeout. China proposed regulations December 2025 to combat AI companion addiction.
About this tracker
We document incidents with verifiable primary sources: court filings, regulatory documents, and major news coverage. This is not speculation or social media claims.
Have documentation of an incident we should include? Contact us.
These harms are preventable.
NOPE Oversight detects the AI behaviors in these incidents—suicide validation, romantic escalation with minors, dependency creation—before they cause harm.