arXiv AI
Relevance as a Vulnerability: How Web Retrieval Degrades Safety Alignment in LLM Agents
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv Machine Learning
Jun 2
Dialectics of Alignment: Harnessing Unsafe Knowledge for Dynamic Safety Routing
arXiv:2606. 00686v1 Announce Type: new Abstract: The prevailing paradigm in large language model (LLM) alignment operates via erasure, filtering unsafe data or training models to strictly refuse harmful prompts.
arXiv AI
Aug 6
Breadcrumbing Search Agents
arXiv:2608. 04565v1 Announce Type: cross Abstract: LLM-based search agents are widely used for information-seeking tasks, but their reliance on external tool returns introduces a critical security risk: web content retrieved during execution is untrusted, exposing agents to prompt injection and goal hijacking.
arXiv AI
Aug 18
The Synthetic Web: Adversarially-Curated Mini-Internets for Diagnosing Epistemic Weaknesses of Language Agents
arXiv:2603. 00801v2 Announce Type: replace Abstract: Language agents increasingly act as web-enabled systems that search, browse, and synthesize information from diverse sources.
arXiv Computation and Language
Sep 11
RAG-Safety-Bench: Reliable Evaluation of Retrieval-Augmented LLM Safety
RAG-Safety-Bench is a benchmark designed to evaluate how retrieval-augmented generation (RAG) affects the safety of large language models (LLMs). It isolates safety impacts by testing four conditions: non-RAG, RAG with an oracle document, RAG with related but non-answer documents, and RAG with random safe documents. Results on five open-source LLMs reveal an inverse relationship between benign and unsafe capabilities, show that baseline safety guardrails do not guarantee safety in RAG, and confirm that even benign documents can trigger unsafe generation.
arXiv Computation and Language
Oct 19
Disentangling Language Effects on Tool Admission and Post-Retrieval Safety
arXiv:2610.23093v1 Announce Type: new Abstract: LLM agents increasingly use retrieval tools across languages, yet their safety evaluation remains English-centric. It remains unclear whether language...
Hugging Face Trending Papers
Sep 10
RAG-Safety-Bench: Reliable Evaluation of Retrieval-Augmented LLM Safety
RAG-Safety-Bench is a benchmark designed to evaluate how retrieval-augmented generation (RAG) affects the safety of large language models (LLMs). It isolates safety impacts by testing four conditions: non-RAG, RAG with an oracle document, RAG with related but non-answer documents, and RAG with random safe documents. Results on five open-source LLMs reveal an inverse relationship between benign and unsafe capabilities, show that standard safety guardrails do not guarantee safety in RAG, and confirm that even benign documents can trigger unsafe outputs.