document-poisoning

1 article

sort: new top best

bug-bounty529 xss292 rce162 google143 account-takeover122 bragging-post118 facebook107 exploit105 privilege-escalation102 microsoft95 authentication-bypass94 open-source94 malware92 csrf88 cve76 stored-xss75 access-control75 ai-agents66 web-security65 reflected-xss63 phishing60 writeup57 input-validation52 sql-injection52 information-disclosure51 ssrf51 cross-site-scripting49 reverse-engineering49 smart-contract49 api-security48 defi48 apple47 tool47 privacy47 ethereum45 vulnerability-disclosure42 web-application40 ai-security39 opinion38 responsible-disclosure37 llm37 burp-suite37 browser37 web337 automation36 race-condition36 remote-code-execution35 lfi34 dos34 credential-theft34

0 7/10

Document poisoning in RAG systems: How attackers corrupt AI's sources

research

Demonstrates document poisoning attacks against RAG systems where malicious documents injected into vector databases can manipulate LLM outputs with 95% success rate on small corpora, and evaluates five defense layers including embedding anomaly detection which reduces attack success to 20% standalone.

rag-security document-poisoning prompt-injection ai-security embedding-attacks retrieval-augmented-generation defense-mechanisms anomaly-detection chromadb local-llm

LM Studio Qwen2.5-7B-Instruct ChromaDB mcp-attack-labs PoisonedRAG

aminrj.com · aminerj · 2 days ago · details · hn