threatdeck
Live 14:45 UTC ·

AI Stack Alert High TD-2026-0022 For CISO, SOC / IR, AppSec, AI Eng AI-assisted

Google Mandiant warns LLMJacking fraud reached $1M per victim in enterprise incidents

Google's Mandiant team reports responding to multiple LLMJacking incidents this year across financial services, retail, technology, and transportation. Attackers use stolen LLM API keys or cloud identity tokens to run model workloads, and some organizations saw cumulative fraudulent charges reach $1m before access was revoked.

Why it matters

Stolen cloud credentials can turn into rapid, large billing losses that are charged to the victim before detection. Google expects criminals to prioritize this over cryptojacking.

Do this

  1. Inventory all LLM API keys, cloud IAM credentials, service account keys, and GitHub personal access tokens that can reach model endpoints.
  2. Rotate keys and tokens that are long-lived, unused, or shared, and enforce short-lived credentials where possible.
  3. Set cloud billing alerts and spend quotas on AI model services so that sudden cost spikes trigger fast review.
  4. Monitor for anomalous model invocation volume, new regions, and GPU provisioning events in cloud audit logs.
  5. Review LLM routing and gateway middleware for patch status and restrict its exposure.

Signals

  • NoCISA KEV
  • 34/100Priority
  • 70%Confidence

Evidence

“Across individual incidents, we observed fraudulent charges associated with resource consumption exceeding $100,000 in a single day, with cumulative charges at some organizations reaching $1m before access was revoked.”

Infosecurity Magazine

“Threat actors bypass front-end chatbots to target backend model resources, usually via LLM API keys or exposed or stolen cloud identity tokens”

Infosecurity Magazine

Sources

Affected / related

Google Mandiant, LLMJacking, LLM Mixers

Published 2026-11-03 10:45 UTC · updated 2026-11-10 00:24 UTC · AI-assisted summary, checked against quoted sources · How we score

Following

Follow the vendors, products, AI tools, techniques and actors you care about. Matching stories, CVEs and exploited bugs are marked, a Following tab appears in the feed, and returning visits show what changed. Saved only in this browser.

Presets

Get updates

A personal RSS feed of new stories on everything you follow. Paste it into any feed reader, Slack or Teams RSS app.