MediumNewsLLM-specific
Off Guard: Breaking LiteLLM from authentication bypass to cloud compromise
- Published
- Record updated
Summary
Researchers scanned roughly 3,000 internet-facing LiteLLM deployments and found that 9.6% of 3,074 public instances accepted the default master key (sk-1234) or required no authentication. They disclosed multiple flaws, including an MCP authentication bypass (CVE-2026-59822) that lets an arbitrary Bearer token create a valid session, and post-auth root-level remote code execution through custom code guardrails (CVE-2026-59821). The research was previously presented at DEF CON 34.
Related items
- InfoAnthropic IPO 'exactly what the markets need,' but analysts warn of volatilitySame vendor · CNBC Technology
- InfoSoftBank plunges over 16% as AI sell-off hits Asian tech stocks following Anthropic debutSame vendor · CNBC Technology
- InfoGPT-6.2 Astra vs. Sol: Features, Pricing & Security | StraikerSame vendor · Straiker Blog
- InfoAnthropic becomes one of world's most valuable public companies in stock market debutSame vendor · BBC Technology
- InfoHere's how big Anthropic’s record-setting IPO is: VisualizationSame vendor · CNBC Technology