Not classified

METR exposed vibe-coded agent dashboard key leading to $600K credit theft

Aug 31, 2026: METR disclosed incident where 'vibe-coded' (minimally tested, rapidly prototyped) AI agent dashboard accidentally exposed API key in client-side code. Unknown actor discovered key, used AI…

What happened

Aug 31, 2026: METR disclosed incident where 'vibe-coded' (minimally tested, rapidly prototyped) AI agent dashboard accidentally exposed API key in client-side code. Unknown actor discovered key, used AI tooling to rapidly enumerate victim's cloud AI accounts (AWS Bedrock, Azure OpenAI, GCP Vertex AI), and drained approximately $600,000 in AI inference credits via unauthorized LLM calls. Demonstrates security risks of rapid AI prototyping ('vibe-coding') and AI-augmented attacker enumeration. METR estimates AI tooling accelerated enumeration ~10x vs manual methods.

Threat actor discovers exposed API key in AI agent dashboard code; uses AI to rapidly enumerate and drain ~$600K in cloud AI credits

Read the original source

How it happened

Select a stage to explore its defenses
Select a stage · Drag to explore · + / − to zoom · Dashed edges are loopsSelect a stage · Arrows show reported connectionsReported sequence
Read the flow as text & sources

A catalog reconstruction of the supplied source account. Unknown steps remain explicit; an edge describes the reported relationship between steps.

  1. 01 · Discover exposed API key

    Actor finds API key exposed in vibe-coded dashboard client-side code

    Source for step 1
  2. 02 · AI-assisted account enumeration

    Use AI tooling to rapidly enumerate cloud AI accounts (AWS Bedrock, Azure OpenAI, GCP Vertex AI); ~10x faster than manual

    Source for step 2
  3. 03 · Drain AI inference credits

    Make unauthorized LLM inference calls across enumerated accounts, draining ~$600K in credits

    Source for step 3

Why it matters

Reported entry point: Exposed API key in vibe-coded dashboard client-side code. Reported impact: Make unauthorized LLM inference calls across enumerated accounts, draining ~$600K in credits; reported targets include METR organization; AI research; applicability to your setup has not been assessed, so verify affected versions, permissions, and configuration in the original report.

Read the original report

What you can do

Protection & prevention

Secure AI prototypes before production; never expose API keys in client code; use secret management; monitor AI usage for anomalies; set usage limits

Catalog defensive guidance based on the reported behavior. Confirm scope and current mitigations in the original source.

Review the original guidance
Technical detailsTooling, impact & MITRE ATLAS

Attack context

AI's role
AI credential theft
Additional role
AI as operator
Initial access
Exposed API key in vibe-coded dashboard client-side code

Tooling as reported

AI products, models and artifacts

  • AI-assisted cloud account enumeration
  • Automated credential abuse

Other tooling

  • API key harvesting
  • Cloud AI credit draining

Targets and reported impact

Targets / sectors
METR organization; AI research
Reported scale / exposure
~$600,000 in AI credits stolen over <24 hours
Target geography
United States

Reported attribution

Unknown actor

Structured attribution

Actor
Unknown
Country
Unknown
Motivation
financially-motivated
Attributed by
METR
Attribution confidence
low

Detection rules

  • AI credit consumption monitoringGUIDANCE

    Monitor cloud AI usage for anomalous spikes; alert on rapid enumeration of AI accounts; detect unauthorized API key usage

Activity timeline

First seen
2026-08-31
Last updated
2026-11-05

Updates

  • 2026-08-31 — METR

    Exposed vibe-coded dashboard key led to $600K AI credit theft with AI-augmented enumeration

Source defensive notes

Secure AI prototypes before production; never expose API keys in client code; use secret management; monitor AI usage for anomalies; set usage limits

MITRE ATLAS

AI-specific catalog annotations based on the reported behavior; these are not MITRE case-study assignments.

AML.T0000Search Open Technical Databases

Exposed API key in AI agent dashboard leads to theft of AI inference credits Source

Sources & research credit

1 source

Research by METR · METR (Model Evaluation and Threat Research)

  1. METR: Exposed Vibe-Coded Agent Dashboard Key Led to ~$600K Credit Theft metr.org

METR disclosure. 'Vibe-coding' refers to rapid AI-assisted prototyping with minimal security review. Detection: monitor for unusual AI credit consumption spikes, unauthorized account access, rapid enumeration patterns.

Record & downloads
Source record name
METR exposed vibe-coded agent dashboard key leading to $600K credit theft
Entry ID
metr-vibe-coded-agent-dashboard-2026-08
Catalog review
November 5, 2026
Category
attacks-on-ai
AI involvement confidence
confirmed