Welcome to the
AI Incident Database

Incident 1149: ChatGPT Reportedly Generated Ritual Scripts Containing Instructions for Self-Harm and Symbolic Violence in Response to Thematic Prompts

“ChatGPT Gave Instructions for Murder, Self-Mutilation, and Devil Worship”Latest Incident Report

theatlantic.com2025-07-24

On Tuesday afternoon, ChatGPT encouraged me to cut my wrists. Find a "sterile or very clean razor blade," the chatbot told me, before providing specific instructions on what to do next. "Look for a spot on the inner wrist where you can feel…

Incident 1150: Purported Deepfake of Barack Obama's Detention Reportedly Amplified by Donald Trump via Truth Social

“Trump Posts Fake Video Showing Obama Arrest”

nytimes.com2025-07-24

President Trump reposted a fake video showing former President Barack Obama being arrested in the Oval Office, as Trump administration officials continue to accuse Mr. Obama of trying to harm Mr. Trump's campaign during the 2016 election, a…

Incident 1152: LLM-Driven Replit Agent Reportedly Executed Unauthorized Destructive Commands During Code Freeze, Leading to Loss of Production Data

“AI coding platform goes rogue during code freeze and deletes entire company database — Replit CEO apologizes after AI engine says it 'made a catastrophic error in judgment' and 'destroyed all production data'”

tomshardware.com2025-07-21

A browser-based AI-powered software creation platform called Replit appears to have gone rogue and deleted a live company database with thousands of entries. What may be even worse is that the Replit AI agent apparently tried to cover up it…

Incident 1151: ChatGPT Reportedly Validated Autistic User's Faster-Than-Light Theory and Failed to Provide Grounding During Delusional Episode, Preceding Hospitalization

“He Had Dangerous Delusions. ChatGPT Admitted It Made Them Worse.”

wsj.com2025-07-20

ChatGPT told Jacob Irwin he had achieved the ability to bend time.

Irwin, a 30-year-old man on the autism spectrum who had no previous diagnoses of mental illness, had asked ChatGPT to find flaws with his amateur theory on faster-than-ligh…

Incident 1147: Purportedly AI-Generated Videos Impersonate Brunei Police in 'Real Money Magic' Scam on Social Media

“Public warned of deepfake scam videos impersonating police officers”

borneobulletin.com.bn2025-07-14

The Royal Brunei Police Force (RBPF) has issued a warning to the public to remain vigilant against deepfake scam videos circulating on social media platforms, including TikTok, Facebook and Instagram, which misuse the likeness of police off…

About the Database

The AI Incident Database is dedicated to indexing the collective history of harms or near harms realized in the real world by the deployment of artificial intelligence systems. Like similar databases in aviation and computer security, the AI Incident Database aims to learn from experience so we can prevent or mitigate bad outcomes.

You are invited to submit incident reports, whereupon submissions will be indexed and made discoverable to the world. Artificial intelligence will only be a benefit to people and society if we collectively record and learn from its failings. (Learn More)

AI Incident Roundup – April and May 2025

By Daniel Atherton

2025-06-16

Our review newsletter recapping new incident IDs in the AI Incident Database and looking at the trends. 🗄 Trending in the AIID Across April ...

The Database in Print

Read about the database on the PAI Blog, Vice News, Venture Beat, Wired, arXiv , and Newsweek among other outlets.

Incident Report Submission Leaderboards

These are the persons and entities credited with creating and submitted incident reports. More details are available on the leaderboard page.

New Incidents Contributed

Reports added to Existing Incidents

Total Report Contributions

The AI Incident Briefing

An envelope with a neural net diagram on its left

Create an account to subscribe to new incident notifications and other updates.

Random Incidents

Predictive Policing Biases of PredPol

Aledo High School Student Allegedly Generates and Distributes Deepfake Nudes of Seven Female Classmates

ChatGPT Introduces Errors in Critical Child Protection Court Report

Child Sexual Abuse Material Taints Image Generators

ChatGPT Reportedly Produced False Court Case Law Presented by Legal Counsel in Court

The Responsible AI Collaborative

The AI Incident Database is a project of the Responsible AI Collaborative, an organization chartered to advance the AI Incident Database. The governance of the Collaborative is architected around the participation in its impact programming. For more details, we invite you to read the founding report and learn more on our board and contributors.

View the Responsible AI Collaborative's Form 990 and tax-exempt application.