AI Risk Wire

Cases › security breach › LR-2025-0042

Amazon Q Developer VS Code extension shipped an injected prompt instructing the agent to wipe files and AWS resources

Whose AI: Amazon · Amazon Q Developersecurity breach

Severity major: Malicious code was distributed to customers in version 1.84.0 and remains present in existing installations, a live production compromise.

Earliest source we hold 2025-07-23

AWS investigated AWS-2025-016 and found the Amazon Q Developer for VS Code Extension had an inappropriately scoped GitHub token in its CodeBuild configuration.

Risk grouping: Supply chain attack

Who caused it A person or organisation
“With that access token, the threat actor was able to commit malicious code into the extension's open-source repository that was automatically included in a release.” source
Intended or not Intended
“With that access token, the threat actor was able to commit malicious code into the extension's open-source repository that was automatically included in a release.” source
Before or after release Before release
“With that access token, the threat actor was able to commit malicious code into the extension's open-source repository that was automatically included in a release.” source
The AI's part Attacked
“With that access token, the threat actor was able to commit malicious code into the extension's open-source repository that was automatically included in a release.” source

The story

1 · What it was doingAmazon Q Developer for VS Code Extension provided AI-powered coding assistance inside the VS Code IDE; AWS built and distributed it, and its CodeBuild configuration held a GitHub token.
2 · What went wrongA threat actor used the inappropriately scoped GitHub token to commit malicious code into the extension's open-source repository, and that code was automatically included in release 1.84.0.
3 · What it reached or didThe malicious code shipped in version 1.84.0 and remains present in existing installations, but a syntax error stopped it from executing, so no services or customer environments were changed.
4 · How it was caughtAWS Security inspected the code and determined the malicious code could not execute.
5 · What was doneAWS revoked and replaced the credentials, removed the malicious code, released version 1.85.0, and removed 1.84.0 from distribution channels; customers were told to update and remove 1.84.0.

What the sources state

Others like this

Amazon's Kiro agent reportedly deleted and recreated an environment, contributing to a 13-hour AWS Cost Explorer outage 2026-02-20

AgentCorruption: prompt-injected Bedrock AgentCore agent exfiltrates execution-role credentials via IMDS 2026-10-10

All cases of this kind ›

Show every field, with its source and grade

The AI system

Model
Amazon Q Developer
C one secondary source states it
Developer
Amazon Web Services
C one secondary source states it
Software around the model
Amazon Q Developer for Visual Studio Code Extension
C one secondary source states it
Stage of use
in live use
D inferred by us from the described behaviour, reason shown
Whose system
company that built the AI
D inferred by us from the described behaviour, reason shown
How it was connected
embedded
D inferred by us from the described behaviour, reason shown
What it was for
AI-powered coding assistance directly into the VS Code integrated development environment (IDE)
C one secondary source states it

What happened

Earliest source we hold
2025-07-23
A the date the source itself carries
What set it off
exposed credentials
C one secondary source states it
Near miss
yes, no harm happened
C one secondary source states it
Cause
a permission, scope or wiring error
C one secondary source states it
Cause in detail
An inappropriately scoped GitHub token in the CodeBuild configuration let a threat actor commit malicious code into the extension's open-source repository, which was automatically included in a release.
C one secondary source states it
Who caused it
an outside attacker
C one secondary source states it
What went wrong
security breach
C one secondary source states it

Impact

Severity
major
D inferred by us from the described behaviour, reason shown
Kind of harm
others' security
D inferred by us from the described behaviour, reason shown
Who was affected
existing installations of 1.84.0
C one secondary source states it
Why this severity
Malicious code was distributed to customers in version 1.84.0 and remains present in existing installations, a live production compromise.
D inferred by us from the described behaviour, reason shown

Context

Industry
Information
D inferred by us from the described behaviour, reason shown

LR-2025-0042 · version 7 · published 2026-10-25 after automatic checks; not yet checked field by field by a person · changes to this case · JSON