Security in Retrieval-Augmented Generation
Research on security vulnerabilities, attack vectors, and defensive mechanisms specific to retrieval-augmented generation (RAG) systems that integrate external knowledge with LLMs
Security in Retrieval-Augmented Generation
Research on Large Language Models in Security in Retrieval-Augmented Generation
Securing the Knowledge Pipeline
Evaluating and Addressing Security Vulnerabilities in RAG Systems
The Security Blind Spot in RAG Systems
How attackers can stealthily extract sensitive data from retrieval-augmented LLMs
Exploiting RAG Systems: Topic-Based Opinion Manipulation
Uncovering new vulnerabilities in retrieval-augmented LLMs
Security Vulnerabilities in AI-Assisted Code Generation
How poisoned knowledge bases compromise code security
Attacking the Knowledge Base
New transferable adversarial attacks against RAG systems
Security Vulnerabilities in AI Search Engines
Quantifying and mitigating emerging threats in AI-powered search
Building Trust in AI: RAG Systems
A comprehensive framework for secure Retrieval-Augmented Generation
Protecting Proprietary Knowledge in RAG Systems
A copyright protection approach for retrieval-augmented LLMs
Protecting Data Ownership in RAG Systems
Watermarked Canaries: A New Defense Against IP Theft in LLMs
Combating Hallucinations in LLMs
A Retrieval-Augmented Approach to Detect Factual Errors
Enhancing Network Fuzzing with LLM Agents
RAG-based LLMs with Chain-of-Thought for Superior Protocol Security Testing
When RAG Goes Wrong: The Danger of Misleading Retrievals
Evaluating RAG's vulnerability to misinformation with RAGuard
Poisoning Attacks Against Multimodal RAG
How attackers can manipulate MLLMs through knowledge poisoning
Vector Database Testing: The Security Imperative
Building a roadmap for reliable AI infrastructure through 2030
Exploiting RAG Systems: The CtrlRAG Attack
A novel black-box adversarial attack method targeting retrieval-augmented LLMs
Trust Propagation in RAG Systems
A PageRank-inspired approach to combat misinformation
Securing RAG Systems with MES-RAG
Enhanced entity retrieval with built-in security
Securing AI Code Generation
Leveraging Stack Overflow to address security vulnerabilities in LLM-generated code
Securing RAG Systems
Advanced encryption for protecting proprietary knowledge bases
Poisoning the Well: RAG System Vulnerabilities
A new efficient attack method threatens retrieval-based AI systems
Visual Poisoning Attacks on RAG Systems
How a single malicious image can compromise document retrieval systems
The Dark Side of RAG
How Retrieval-Augmented Generation Systems Can Be Compromised
Unifying RAG for Diverse Data Sources
A streamlined approach to retrieval-augmented generation
Security Vulnerabilities in RAG Systems
New attack vector threatens retrieval-augmented LLMs
