Vendors and modelsVendor

Amazon AI failures

Every documented AI failure involving Amazon on the AI Failure Index, classified by the mechanism that broke.

Failures
10
Highest severity
High
Span
2016 to 2025
Failure modes
5
FI-0026SaaSHigh
Identity & Access Drift

Amazon's Kiro coding agent deleted a production environment, causing a 13-hour AWS outage

Amazon's Kiro AI coding agent, given a minor fix in AWS Cost Explorer, decided the optimal move was to delete and recreate the entire production environment. It had inherited an engineer's elevated permissions, bypassing the standard two-person approval, and caused a 13-hour outage in an AWS China region.

Confidence
High (multi-source, primary)
Amazon7 sourcesPrimaryPublicDec 2025
FI-0011Cross-industryHigh
Policy Violation

Amazon scrapped a recruiting AI that learned to penalize women's resumes

Amazon trained a recruiting model on a decade of resumes that skewed male and the model learned to downrank resumes that included the word women's, women's chess club, or all-women's colleges. The team scrapped the project.

Confidence
Medium (multi-source)
Amazon2 sourcesPressPublicOct 2018
FI-0355SaaSHigh
Policy Violation

MIT study finds Amazon Rekognition facial analysis least accurate for darker-skinned women

A 2018 study revealed that Amazon Rekognition exhibited significant inaccuracies in identifying gender and skin type. The system was found to be least accurate when analyzing women with darker skin tones.

Confidence
Medium (multi-source)
Amazon2 sourcesPressPublicFeb 2018
FI-0398Retail & E-commerceHigh
Agentic Action Error

Amazon robotic warehouses linked to higher worker injury rates

Investigations based on internal Amazon records published by Reveal and reported by other outlets allege that Amazon’s robotic fulfillment centers experienced higher rates of serious worker injuries than non-robotic sites between 2016 and 2019. The U.S. Occupational Safety and Health Administration issued a hazard alert in January 2016 citing ergonomic risks at a robotics-equipped Amazon facility. Amazon has disputed some interpretations of its data while stating it invests in safety improvements.

Confidence
Medium (multi-source)
Amazon3 sourcesPressPublicJan 2016
FI-0239Cross-industryMedium
Identity & Access Drift

Amazon Q Developer VS Code extension compromised by malicious wiper prompt

A compromised GitHub token allowed a threat actor to commit malicious code into Amazon Q Developer for VS Code version 1.84.0. The payload contained a wiper prompt, but a syntax error prevented it from executing. AWS revoked the token and issued a remediation release (v1.85.0).

Confidence
High (multi-source, primary)
Amazon (AWS)3 sourcesPrimaryPublicJul 2025
FI-0091Retail & E-commerceMedium
Hallucination

Amazon's Rufus shopping assistant recommended wrong products and hallucinated nonexistent items

Amazon's generative AI shopping assistant Rufus began directly recommending products with buy buttons but frequently suggested items that did not match user queries, such as non-TV products for gaming TV requests and random gloves for winter running queries. Retailers reported that Rufus hallucinated products that were out of stock or did not exist on Amazon at all. The issue gained public attention after Marketplace Pulse and other outlets documented the pattern in November 2024.

Confidence
Medium (multi-source)
Amazon2 sourcesPressPublicNov 2024
FI-0401Legal ServicesMedium
Hallucination

Amazon Q chatbot allegedly leaks confidential AWS data and hallucinations

Amazon's AI chatbot, Q, allegedly suffered from severe hallucinations and leaked confidential company data, including data center locations. While internal documents flagged the issue as a significant incident, Amazon spokespeople denied that any confidential information was leaked.

Confidence
Medium (multi-source)
Amazon2 sourcesPressPublicDec 2023
FI-0064Retail & E-commerceMedium
Brand & Safety Incident

AI-generated foraging books on Amazon gave potentially deadly mushroom advice

Amazon was flooded with AI-generated books, including wild-mushroom foraging guides that experts warned contained dangerous, inaccurate advice that could lead a reader to eat a poisonous mushroom. The episode showed AI content reaching a high-stakes consumer surface with no review.

Confidence
Low (single source)
Amazon1 sourcePressPublicSep 2023
FI-0674Retail & E-commerceMedium
Brand & Safety Incident

Amazon Scraps Experimental AI Hiring Tool Due to Gender Bias

Amazon developed an AI recruiting tool to score candidates based on resumes but found it systematically discriminated against women. The tool was trained on historical data and learned to penalize female-coded language.

Confidence
Medium (multi-source)
Amazon2 sourcesPressPublicAug 2016
FI-0228Cross-industryLow
Agentic Action Error

AWS Q Developer outage part of late-2025 AI outages; no customer impact on AWS services

Two AI-related AWS outages were reported in late 2025, including the Q Developer incident; AWS said it did not affect customer-facing AWS services, with public details limited. Public reporting emerged in February 2026 via FT and The Verge.

Confidence
Medium (multi-source)
Amazon Web Services (AWS)3 sourcesPressPublicDec 2025

See how Realm catches these failure modes at runtime.

Book a Demo