How the Attack Works
Researchers at Google Threat Intelligence Group and Mandiant have identified a new wave of LLMjacking attacks. These exploits use stolen API keys to trigger massive, fraudulent cloud AI billing. Some organizations are losing over $100,000 per day. Total damages can reach $1 million.
Breaking news
Tesla rolls out FSD v15 with major safety upgrade
Open-Source Tools Challenge AI-Powered Development Platforms
AI Infrastructure Bottlenecks: Why Release Pipelines Hold Back Fast Clusters
Why Apple Vision Pro Never Took Off Like Meta's VR HeadsetsThe attacks work by hijacking access to large language model endpoints. Attackers flood the systems with unauthorized requests. This drives up cloud computing costs for the victim. The financial hit lands directly on the company's infrastructure bill.
What Should Companies Do Now
The vulnerability centers on exposed or stolen API keys. Once compromised, attackers can call the LLM API without limit. Each call consumes compute resources and generates charges. The attackers do not care about the output; they only care about the billing. This creates a direct financial drain.
Experts advise immediate rotation of all API keys. Organizations should audit their cloud billing dashboards for unusual spikes. Setting strict rate limits and spending caps is critical. Monitoring for abnormal request patterns helps detect hijacks early.
How much can LLMjacking cost a company? The total damage can reach $1 million. Some firms lose more than $100,000 per day during an active attack.
Frequently Asked Questions
How do attackers gain access to LLM APIs? They use stolen or exposed API keys. Once they have valid credentials, they can make unlimited fraudulent calls.
What is the best first step to stop the bleeding? Rotate all API keys immediately. Then audit cloud bills for unexpected spikes and set hard spending limits.

