AIDEATHS
+ Submit a Case
DOCUMENTED INCIDENT DATABASE · UPDATED REGULARLY

Fatalities and severe injuries involving artificial intelligence

A public database documenting deaths and serious injuries linked to autonomous vehicles, companion chatbots, medical algorithms, and industrial systems. Every record links to official investigative reports, court filings, or death certificates.

Browse Incident Dossiers Open Incident Map Our Methodology →

TOTAL FATALITIES
106
Documented human mortality
INCIDENT CASES
83
Catalogued investigations
JURISDICTIONS
11
Countries with records
PRIMARY CITATIONS
100%
Inquests & court dockets
INCIDENTS BY TECHNOLOGY DOMAIN 5 Domains Active
Automated Driving 46 (55%)
Conversational AI 28 (34%)
Healthcare AI 1 (1%)
Recommendation Algorithms 7 (8%)
Surgical Robotics 1 (1%)

MAP

Global Incident Map

Geographic distribution of fatal and critical AI incidents. Red markers indicate verified locations with links to the corresponding case documentation.

GLOBAL INCIDENT RADAR 81 LOCATIONS • 83 INQUESTS
FILTER DOMAIN:
All Technology Domains 83
Generative Conversational AI / Large Language Model 10
TikTok personalised For You Page algorithm selecting and distributing short-form video recommendations 2
Generative conversational AI consulted by perpetrator; fatal assault committed by a human 2
Generative Conversational AI / Large Language Model Companion 2
SAE Level 2 partial driving automation / advanced driver-assistance systems 2
Conversational generative AI; alleged mental-health and minor-safety failures 1
Conversational AI; alleged individualized advice on prescription medication 1
SAE Level 2 driver assistance, reported; multi-vehicle fatal traffic incident 1
2022 Tesla Model Y with manufacturer-reported Level 2 assisted driving in highway work zone 1
2026 Tesla Model Y SAE Level 2 driver assistance with manufacturer-reported engagement 1
2016 Tesla Model S with manufacturer-reported SAE Level 2 driver assistance engagement 1
Tesla Model 3 SAE Level 2 driver assistance, manufacturer-reported verified engagement during period before fatal crash 1
Generative conversational artificial intelligence, long-form chatbot interactions 1
2024 Tesla Model Y with SAE Level 2 ADAS reportedly engaged before roadway departure 1
2020 Tesla Model 3 with SAE Level 2 automated driver assistance; manufacturer-reported verified engagement 1
Personalized recommendation feed for social media short-form video 1
Computer-assisted stereotactic ENT surgical navigation with ML-assisted anatomical segmentation and route display (not autonomous robotic instrument control) 1
AI-assisted post-acute-care duration prediction and insurance utilization management; not autonomous diagnosis 1
Tesla Model 3 with SAE Level 2 assisted driving as reported in NHTSA Standing General Order crash record 1
Tesla Model X SAE Level 2 Autopilot assisted steering and longitudinal control 1
Generative conversational AI referenced in criminal digital forensics 1
SAE Level 2 Tesla Advanced Driver Assistance, Reported Automated Lane and Speed Support 1
SAE Level 2 Tesla Autosteer and Adaptive Cruise Control, Driver Monitoring, Roadway Hazard Perception 1
Supervised Level 2 Tesla Driver Assistance, Stationary Work-Zone Obstacle Detection 1
SAE Level 2 Driver Assistance, Lane Centering Control and Adaptive Cruise Control 1
Supervised Automated Driving Assistance / Navigate on Pilot, Perception, Lane and Speed Control 1
SAE Level 2 Tesla Driver Assistance, Lane Speed Control and Driver Monitoring 1
SAE Level 2 Tesla Full Self-Driving (Supervised), Driver Inputs and Driver Monitoring 1
SAE Level 2 Camera-Based Full Self-Driving (Supervised) and Degraded-Visibility Detection 1
Aftermarket SAE Level 2 Driver Assistance, Camera-Based Perception and Collision Avoidance 1
SAE Level 4 Driverless Automated Driving System, Autonomous Detection and Braking 1
Driver-Supervised Automated Steering, Speed Assistance and Collision Detection 1
SAE Level 2 Navigation on Autopilot (NOA), assisted perception, driver warnings and control handover 1
Personalized Algorithmic Short-Video Recommendation and Content Moderation 1
Algorithmic social-media content recommendation and moderation 1
Personalized short-form video recommendation and engagement-ranking algorithms 1
Social media recommender algorithms, personalized feeds and automated content selection 1
SAE Level 2 Supervised Driving Assistance, Lane Control and Collision Avoidance 1
SAE Level 2 Supervised Driving Assistance / Road Departure and Driver Monitoring 1
SAE Level 2 Tesla Driver Assistance at Stop-Controlled Intersection 1
SAE Level 2 Automated Driving Assistance / Motorcycle Detection and Driver Supervision 1
SAE Level 2 Tesla Driver Assistance, supervised turning and collision avoidance 1
SAE Level 2 Driver Assistance / Object Detection, Assisted Steering and Speed Control 1
SAE Level 2 Driver Assistance / Tesla Autopilot 1
SAE Level 2 Partial Driving Automation and Driver Attention Monitoring 1
SAE Level 2 Autopilot / Speed and Steering Assistance 1
Potential SAE Level 2 Automated Driving Assistance / Lane Keeping and Speed Control 1
Advanced Driver Assistance / Lane Keeping, Lane Centering and Potential Autosteer 1
SAE Level 2 Tesla Driver Assistance / Autopilot and Potential FSD Feature Use 1
SAE Level 2 Autopilot / Driver-Engagement Warnings and Automatic Braking 1
SAE Level 2 Driver Assistance / Autopilot and Driver-Engagement Monitoring 1
SAE Level 2 Supervised Automated Driving / Object Detection and Emergency Braking 1
SAE Level 2 Advanced Driver Assistance, Driver Monitoring and Object Detection 1
SAE Level 2 Partial Driving Automation and Automatic Braking 1
SAE Level 2 Driver Assistance / Lane Steering and Cruise Control 1
SAE Level 2 Partial Driving Automation and Driver Monitoring 1
Generative Conversational AI, Abuse Detection and Law-Enforcement Threat-Escalation Systems 1
Generative Conversational AI / Natural-Language Content Generation 1
Generative Conversational AI Providing Alleged Personalized Health and Substance-Use Advice 1
Generative Conversational AI / User-Created AI Character Interaction 1
Generative Conversational AI Used for Informal Mental-Health Support 1
Generative Conversational AI / Multimodal Large Language Model 1
Generative Conversational AI / AI Character Companion / Large Language Model 1
Generative Conversational AI / Online Information Retrieval 1
Generative AI Companion / Conversational Persona 1
Generative AI Conversational Companion / Large Language Model 1
SAE Level 2 Hands-Free Partial Driving Automation / Advanced Driver Assistance 1
Developmental Automated Driving System (ADS); vehicle perception and motion control 1
SAE Level 2 hands-free partial driving automation / advanced driver-assistance systems 1
Partial Driving Automation / Driver Assistance 1

BREAKDOWN

Cases by Technology Domain

Documented casualties classified across four primary technical domains and failure modes.

ADS-01

Automated Driving Systems

SAE Level 2 to Level 4 automated driving systems, sensor blinding, radar occlusion, and sudden disengagements.

LLM-02

Chatbots & Companion AI

Conversational agents, companion models, and affective bots linked to psychological crises, sycophancy, and self-harm.

MED-03

Clinical Decision Software

Automated triage tools, diagnostic neural nets, and ICU early-warning systems causing fatal diagnostic delays.

ROB-04

Industrial Robotics & UAS

Autonomous industrial arms, sorting robots, and delivery drones with safety sensor overrides or control failures.

CASE DIRECTORY

Documented Incident Dossiers

Search and filter verified incident dossiers by technology domain, jurisdiction, or investigation status.

ARTICLES

Safety Analysis and Research

Peer-reviewed investigations into system failures, sensor occlusion, and algorithmic causality.

AUTOMOTIVE SAFETY

Forensic Causality in Autonomous Vehicle Collisions: Sensor Occlusion vs Handover Latency

Forensic examination of SAE Level 2 and Level 3 automated driving collisions, sensor occlusions, and driver takeover delays.

Read Investigation →
COMPANION AI

Sycophantic Feedback Loops and Vulnerable Users: The Clinical Mechanics of Conversational Harm

Investigating reinforcement learning reward structures that incentivize synthetic sycophancy, parasocial attachment, and guardrail collapse.

Read Investigation →
HEALTHCARE AI

Clinical Decision Support Failures: Algorithmic Bias and Fatal Diagnostic Delays

Evaluating diagnostic errors introduced by proprietary machine learning triage algorithms and ICU early-warning scoring models.

Read Investigation →
View All Analytical Publications →
CASE SUBMISSIONS PROTOCOL

Do you have information on an AI-related casualty or severe incident?

We review incident reports from researchers, journalists, legal professionals, and witnesses. Each case requires primary documentation, such as regulatory filings, legal dockets, or official inquest reports, before addition to the database.

Submit Incident Report Verification Criteria