In November 2026, the Google Threat Intelligence Group reported on multiple incidents where threat actors hijacked enterprise access to Large Language Models in a tactic called LLMJacking. These intrusions generated massive fraudulent cloud charges.
The Mechanics of LLMJacking
Mandiant responded to multiple incidents in 2026 in which threat actors hijacked enterprise access to Large Language Models, a tactic known as LLMJacking. These intrusions impacted organizations in North America, South America, and Asia Pacific across the financial services, retail, technology, and transportation sectors.
Unlike the gradual resource consumption associated with cryptojacking, LLMJacking can generate large spikes in fraudulent cloud charges in a matter of hours. Threat actors bypass front-end interfaces to target backend model resources directly, focusing on identifying publicly exposed LLM endpoints or obtaining valid LLM application programming interface keys.
Monetization and LLM Mixers
The Google Threat Intelligence Group identified a division of labor between threat actors establishing initial access to compromised environments and operators managing downstream monetization infrastructure known as LLM Mixers. LLM Mixers function as illicit routing gateways that proxy, aggregate, and resell unauthorized LLM access to underground subscribers.
Across individual incidents, observers noted fraudulent charges associated with resource consumption exceeding $100,000 USD in a single day, with cumulative charges at some organizations reaching $1 million USD before access was revoked. Threat actors subscribed victim accounts to Pay-As-You-Go agreements and launched high-volume inference streams.
Mitigation and Remediation Guidance
To prevent attackers from abusing cloud LLM services, defenders should prioritize Identity and Access Management and supply chain security. Organizations are advised to utilize centralized secrets management platforms and enforce least privilege access controls.
Additional recommendations include automating secret scanning within codebases, performing routine configuration audits, implementing rate limiting, and enforcing phishing-resistant multi-factor authentication across all cloud identity providers and developer platforms.
What to do
- Utilize a centralized secrets management platform to control the management and rotation of sensitive keys and secrets.
- Ensure users and service accounts only possess the permissions necessary for their specific role and required tasks.
- Automate secret scanning within codebases to ensure API keys and secrets added to source code are identified and removed promptly.
- Perform periodic comprehensive configuration audits of Cloud services to identify potential attack paths and unintended exposure of endpoints.
- Implement a WAF like Google's Cloud Armor to enforce rate limiting for individual client connections.
- Enforce phishing-resistant multi-factor authentication across all cloud identity providers, management consoles, and developer platforms.
Key facts and where they come from
- Mandiant responded to multiple 2026 incidents where threat actors hijacked enterprise access to Large Language Models.
Mandiant responded to multiple incidents in 2026 in which threat actors hijacked enterprise access to Large Language Models (LLMs), a tactic known as "LLMJacking."
- Fraudulent charges associated with resource consumption exceeded $100,000 USD in a single day in observed incidents.
observed fraudulent charges associated with resource consumption exceeding $100,000 USD in a single day
- LLM Mixers function as illicit routing gateways that proxy, aggregate, and resell unauthorized LLM access.
LLM Mixers function as illicit routing gateways that proxy, aggregate, and resell unauthorized LLM access to underground subscribers.
Hacking Techniques tracker: how this works, how it is advancing and how to defend — Attacks using or targeting AI
