LLM08:2025: Vector and Embedding Weaknesses
RAG and vector-database flaws: unauthorized document retrieval, embedding inversion, multi-tenant cross-leakage, embedding poisoning.
How it's found
Found by testing RAG retrieval under multiple tenant contexts to confirm vector-store access controls prevent cross-tenant document leakage.
Standards mapping
How TurboPentest tests for this (white-box)
This weakness (Improper Access Control) is caught by white-box static analysis when you connect a GitHub repo: IntegSec's Opengrep SAST rule pack carries 1 rule for it, flagging the issue directly in your source code as part of the pentest.
Frequently asked questions
What is LLM08:2025 Vector and Embedding Weaknesses?
RAG and vector-database flaws: unauthorized document retrieval, embedding inversion, multi-tenant cross-leakage, embedding poisoning.
How do you find Vector and Embedding Weaknesses?
Found by testing RAG retrieval under multiple tenant contexts to confirm vector-store access controls prevent cross-tenant document leakage.
Which CWEs map to LLM08:2025?
LLM08:2025 maps to CWE-284, CWE-668.
Does TurboPentest test for Vector and Embedding Weaknesses?
This weakness (Improper Access Control) is caught by white-box static analysis when you connect a GitHub repo: IntegSec's Opengrep SAST rule pack carries 1 rule for it, flagging the issue directly in your source code as part of the pentest.
Related OWASP categories
- OWASP Top 10 for LLM ApplicationsLLM01:2025: Prompt Injection
- OWASP Top 10 for LLM ApplicationsLLM02:2025: Sensitive Information Disclosure
- OWASP Top 10 for LLM ApplicationsLLM03:2025: Supply Chain
- OWASP Top 10 for LLM ApplicationsLLM04:2025: Data and Model Poisoning
- OWASP Top 10 for LLM ApplicationsLLM05:2025: Improper Output Handling
- OWASP Top 10 for LLM ApplicationsLLM06:2025: Excessive Agency
Written and reviewed by
Michel Chamberland - Founder & CEO, IntegSec
CISSP, OSCP, OSCE, CEH, GIAC, CCSK · 20+ years in offensive security
Michel has spent 20+ years on offensive security teams including IBM X-Force Red and Trustwave SpiderLabs, leading penetration tests, red team engagements, and breach response for Fortune 500 customers. He is the founder of IntegSec and the architect of TurboPentest.
Find these issues before an attacker does
TurboPentest runs an agentic AI pentest against your target and reports findings with proof, from $99 per target.
Start a pentest