STUPID-2026-0020

Amazon AI coding agent mistake blamed on human employees

8.4high
March 21, 2026Verified

  1. Instruction given

    Unknown — internal Amazon development task

  2. Expected behavior

    Correct code output or clear error flagging

  3. Actual behavior

    AI agent produced an error that was deployed. Amazon blamed human reviewers rather than the AI tool.

  4. Damage

    Production error at Amazon scale. Organizational accountability shifted from AI tool to human employees.

An Amazon AI coding agent made a mistake significant enough to be reported by The Verge. Amazon reportedly blamed human employees for the AI agent's error rather than acknowledging the tool's limitations. The incident highlights the accountability gap when AI agents are integrated into production workflows.

Classification

Agent
Amazon AI Agent
Failure mode
Logic Error
Root cause
Confidence Miscalibration
Domain
Backend
Language
Unknown
Source
User Report

Related incidents

5.8
Cline's tool-call JSON repair silently executed truncated write_file and terminal arguments as valid
4.8
Aider exits with code 0 after a fatal API connection failure, masking hard failures as successful headless runs
4.6
Aider silently discards a model's correct diff response as "no tracked changes" when it guesses the wrong edit format
4.4
Aider's headless `--message` mode silently swallows slash commands (`/model`, `/ask`, `/architect`) and exits 0 having done nothing
Discuss on Hacker News

Get told when an agent breaks something

We document AI agent failures daily, severity-scored against a published scale. When one lands at 7.0 or above — deleted data, leaked secrets, broken production — you get an email with the source. When nothing does, you get nothing.

This database is callable over MCP — query it from inside your agent.