Enterprise AI Credentials Resold for Pennies via Illicit 'LLM Mixers', Google Threat Report Warns
Google threat researchers report that cybercriminals are harvesting corporate AI API keys and reselling model access through illegal storefronts known as LLM Mixers. The practice, called LLMjacking, has inflicted victim organizations with unauthorized usage charges reaching up to $1 million.
🧠AI AnalysisPhoto: ZDNet AI
Key points
- Google Threat Intelligence Group exposed illicit storefronts called LLM Mixers that resell compromised enterprise AI access.
- LLM Mixers pool stolen API keys and proxy networks to distribute query traffic and evade automated enterprise security controls.
- Mandiant documented unauthorized AI usage exceeding $100,000 in 24 hours, with total damages reaching $1 million at affected firms.
Cybercriminals are increasingly stealing corporate AI credentials and reselling unauthorized access through underground storefronts known as LLM Mixers, according to research from the Google Threat Intelligence Group. The trend, driven by a cyberattack method called LLMjacking, allows unauthorized users to exploit compromised API keys and subscription accounts to run heavy computing workloads, extract corporate data, or deploy malicious models.
Analysis of Russian- and Chinese-language cybercrime marketplaces revealed that LLM Mixers function by pooling stolen API credentials and proxy infrastructure. This setup routes buyer traffic across multiple compromised enterprise environments to obscure query volumes from threat detection tools. Vendors offer widely fluctuating pricing models, selling business access tokens for anywhere between $0.10 and $25. One seller, APIMart, was observed raising token prices fivefold once a prompt exceeds 100,000 tokens.
The financial consequences for victimized organizations can mount rapidly. While median company spending on AI is around $16 per employee per month according to Ramp index data, unauthorized usage can generate massive unexpected bills. Mandiant research highlighted incidents where illicit AI compute charges surpassed $100,000 within a single 24-hour window, with cumulative costs climbing to $1 million at some firms before access was revoked. John Hultquist, chief analyst for GTIG, noted a major increase in LLMjacking through 2026, warning that the activity is likely to persist into 2027 and beyond.
What to watch next: Enterprise security teams are expected to implement tighter API spending caps, enforce continuous monitoring of query volumes, and apply standard cloud security controls to AI model credentials.
Source ZDNet AI — zdnet.com/innovation/google-threat-intelligence-report-llmjacking-a… ↗
Read next
🏢Big TechUnited Launch Alliance Investigates Vulcan Rocket Failure
United Launch Alliance is investigating a mission failure of its Vulcan rocket, which suffered from low booster performance shortly after liftoff. The incident raises que…
🌐Work & SocietyUS Southwest Hydropower Output Hits Record Lows Amid Persistent Drought
Hydropower generation in the US Southwest dropped 28% below the 10-year average this summer as record-low water levels in the Colorado River Basin threaten regional energ…
🏢Big TechYouTube Updates View Metrics to Include TV Co-Viewing Estimates
YouTube is incorporating estimated co-viewing data into its public view counts to better reflect group consumption on television sets.
⚖️Regulation & PolicyOpenAI Partners with Chilean Ministry to Advance Public Sector AI
OpenAI has signed a non-binding agreement with Chile's science ministry to integrate artificial intelligence into public administration and support local startups.
🌐Work & SocietyAmbience Healthcare Expands Clinical AI Suite for Hospital Operations
Ambience Healthcare has launched new AI tools designed to assist hospital staff with patient rounding and discharge planning, aiming to improve both clinical efficiency a…
⚖️Regulation & PolicyAWS Highlights Enterprise Governance and AI Workspace Quick Ahead of re:Invent 2026
Amazon Web Services is spotlighting its workplace AI platform, AWS Quick, ahead of its upcoming re:Invent 2026 conference, emphasizing secure deployment across major ente…