Independent AI incident database
The searchable record of real-world AI incidents.
Explore documented AI failures, security events, privacy incidents, harmful outcomes and governance breakdowns across companies, AI systems, industries and countries.
Latest recordAII-1001-40493
Quebec Court Fines Man for AI-Generated Fabricated Legal Documents
Confirmed
A man in Quebec submitted AI-generated fabricated legal documents as part of his legal defense in court.
- Date
- 10 October 2026
- Category
- Reliability and hallucination
Drafted by Raidu Incident DetectionView incident ›
7,670
Curated incident records
11,268
Organizations named
118
Countries
7,670
Classified by category
49
Citing a court or regulator
9
Public sources monitored
Counts represent incidents documented in this dataset, not estimates of true incident prevalence.
Explore the Index
Discover documented incidents by category, organization and source.
All incidentsSearch and filter every published record by title, organization, source and date.Explore the database ›Categories10 harm categories, from privacy and surveillance to safety and wellbeing, with secondary categories per record.Browse categories ›Organizations11,268 companies, agencies and institutions named in documented incidents.Browse organizations ›Policy MonitorAI laws, regulations and standards from their official publishers, with our summaries, section notes and the controls each one sets.Browse policies ›Method and standardsHow an incident qualifies, how it is verified, and how it is classified. Every record is drafted by an automated system. A named editor reviews and approves it once its sources are verified.Read the methodology ›DataDataset v2026.09 as CSV and JSON, with field definitions and release notes.View the data page ›
Recently documented
All incidentsOct 2026Quebec Court Fines Man for AI-Generated Fabricated Legal DocumentsView incident ›
Oct 2026Ukraine Ministry of Defence approves Zmiy Droid 12.7 combat robot for deploymentView incident ›
Oct 2026Ukrainian forces shoot down Russian AI-powered Kamikaze drone KlinView incident ›
Oct 2026German Cybersecurity Agency Warns of AI-Driven Vulnerability Discovery RisksView incident ›
Oct 2026AI-Generated Fake Video Spreads Disinformation About Ukrainian Gas StationsView incident ›
Oct 2026Carlsen Verlag sues OpenAI for copyright infringement of "Das NEINhorn"View incident ›
What the data shows
Patterns across the incidents published so far. The dataset grows as research is completed, so these figures describe the Index, not the world.
Incidents by category
Published incidents by editorial category.
Unit: incidents · Source: AI Incident Index v2026.09 · Updated 10 Oct 2026
Incidents by year of occurrence
When the documented events took place.
Unit: incidents · Source: AI Incident Index v2026.09 · Updated 10 Oct 2026
Where documented incidents happened
Country most directly associated with each incident. 118 countries so far.
Incidents1415+1107+2075+Brazil 219, South Korea 178, France 146, Australia 144, Spain 144, Taiwan 138, Japan 104, Canada 91, Turkey 83, Argentina 61, Mexico 61, Romania 61, Israel 60, Greece 59, Vietnam 53, Iran 44, Malaysia 44, Netherlands 44, Poland 43, Ireland 39, Hungary 33, Nigeria 31, Switzerland 27, Colombia 24, Pakistan 24, Egypt 22, South Africa 21, Hong Kong 20, Portugal 18, Thailand 18, Denmark 17, New Zealand 17, Austria 16, Belgium 16, Philippines 16, Norway 14, Serbia 14, Kenya 13, United Arab Emirates 13, Croatia 12, Moldova 11, Slovakia 11, Sweden 11, Cyprus 9, Chile 8, Finland 8, Palestine 7, Uruguay 7, Albania 6, Saudi Arabia 6, Bangladesh 5, Bosnia and Herzegovina 5, Czechia 5, Iraq 5, Slovenia 5, Venezuela 5, Burkina Faso 4, Dominican Republic 4, Ghana 4, Malta 4, Peru 4, Bolivia 3, Honduras 3, Lebanon 3, Lithuania 3, Nepal 3, Angola 2, Azerbaijan 2, Bahrain 2, Cambodia 2, Ecuador 2, Georgia 2, Jamaica 2, Jordan 2, Kuwait 2, Macao 2, Myanmar 2, North Macedonia 2, Reunion 2, Yemen 2, Afghanistan 1, Algeria 1, Armenia 1, Bahamas 1, Bermuda 1, Brunei 1, Cabo Verde 1, Cameroon 1, Costa Rica 1, El Salvador 1, Estonia 1, Ethiopia 1, Guam 1, Jersey 1, Latvia 1, Mongolia 1, Montenegro 1, Namibia 1, Papua New Guinea 1, Paraguay 1, Puerto Rico 1, Sudan 1, Syria 1, Tanzania 1, Vatican City 1, Zimbabwe 1 not drawn
Unit: incidents · Source: AI Incident Index v2026.09 · Updated 10 Oct 2026
Based on incidents currently recorded in the AI Incident Index. This dataset does not represent the total universe of AI incidents.
Research from the Index
Evidence before conclusions.
Every incident is written from public reporting and classified using a consistent, published methodology.
Counts describe what has been documented in this dataset, never an estimate of how often AI incidents occur.
- Documented
- Every published record is a fully written article, not a copied source blurb.
- Sourced
- Each record links the public evidence behind it once its sources are checked, and notes where else the event has been catalogued.
- Reviewed
- A named editor reviews and approves a record once its sources are verified, and the byline links to their profile.
- Classified
- Records carry an organization and, where the evidence supports one, an editorial category.
- Citable
- APA, MLA, BibTeX and plain-text citations on every incident page.
- Correctable
- Anyone, including the organizations named, can submit a correction at no cost.
Help improve the public record.
Know of an incident we should include, or found something that needs correcting?