Public registry · established 2026

The public record of AI failures in production.

Real and anonymized cases of what production AI did when no one was watching, explained at the mechanism layer. Sourced in the open, updated as verified.

Documented failures, cumulativeSep 2009 to Jun 2026
653failures
indexed
11
Industries
8
Failure modes
1681
Sources cited
Browse all 653 failures →
Last updated June 22, 2026Taxonomy v1.0.0Entries 653 verifiedSources 1681 citedLicense CC-BY 4.0Corrections active

AI failures are not edge cases. They are the steady state of systems shipped at scale. A single incident can be catastrophic, and the only way to prevent the next one is to know what the last one looked like.

653
Documented failures
39
Catastrophic severity
380
High or catastrophic severity
224
Under active regulatory exposure

The eight failure modes we track

Every entry is classified by the mechanism that broke inside the model, not just the outcome. That is the part no other tracker explains.

Failures by AI surface

Recently indexed

The twenty newest records on the index. Every record is classified by the mechanism that broke inside the model, not just the outcome.

IDIncidentModeSeverityIndexed
FI-0700Gaggle surveillance alert leads to arrest and detention of 13-year-old studentBrand & Safety IncidentHighJun 2026
FI-0699KBS AI subtitles broadcast profanity during Artemis II launch livestreamHallucinationLowJun 2026
FI-0697Insurance AI photo-based repair estimates allegedly inaccurateBrand & Safety IncidentMediumJun 2026
FI-0696Uber used Greyball software to deceive law enforcement officialsPolicy ViolationHighJun 2026
FI-0695Spamouflage Dragon uses GAN-generated faces for propaganda accountsTool MisuseHighJun 2026
FI-0694Honda Collision Mitigation Braking System false positives allegedly cause accidentsAgentic Action ErrorHighJun 2026
FI-0693Remoteli GPT-3 Twitter Bot Hijacked via Prompt InjectionPrompt InjectionLowJun 2026
FI-0692Cruise self-driving car injures multiple people in San Francisco collisionAgentic Action ErrorHighJun 2026
FI-0691Toronto Public Health AI water quality model produces false negatives for beach closuresBrand & Safety IncidentHighJun 2026
FI-0690SF SPCA robot patrol used to ward off homeless peopleBrand & Safety IncidentMediumJun 2026
FI-0689Manufacturing robot failure causes worker's death at ASAL Chakan plantAgentic Action ErrorCatastrophicJun 2026
FI-0688Ars Technica Retracts Article After Using AI-Generated Fake QuotesHallucinationMediumJun 2026
FI-0687ChatGPT and Perplexity AI Manipulated to Produce Explicit ContentPrompt InjectionMediumJun 2026
FI-0686Healthcare allocation algorithm reduces essential care hours in multiple US statesBrand & Safety IncidentCatastrophicJun 2026
FI-0684Facebook Ad Moderation AI Fails to Detect Violent Hate SpeechBrand & Safety IncidentHighJun 2026
FI-0683YouTube recommendations pushed 2020 election fraud content to skeptical usersBrand & Safety IncidentHighJun 2026
FI-0682AI Chatbots Provide Inaccurate UK Financial and ISA GuidanceHallucinationHighJun 2026
FI-0681Google AI Overviews and ChatGPT Surface Fraudulent Cruise Hotline ScamHallucinationMediumJun 2026
FI-0680Navy Federal Credit Union facing allegations of racial bias in automated mortgage underwritingPolicy ViolationHighJun 2026
FI-0678Uber Real-Time ID Check deactivates transgender drivers for appearance changesBrand & Safety IncidentMediumJun 2026

Browse all 653 documented failures