PARSE: Provenance-Aware Retrieval Sanitization for Professional Domain LLM Agents
TL;DR
This work introduces PARSE (Provenance-Aware Retrieval Sanitization), a domain-aware, fact-preserving sanitization pipeline that classifies each sentence by injection likelihood, extracts structured facts before rewriting, and verifies fact preservation via a consistency-checking loop.
Abstract
Prompt injection defenses evaluated on synthetic benchmarks do not generalize to real enterprise documents, which are longer, denser, and interleave legitimate authority language with factual content. We demonstrate this gap with a benchmark of 122 tasks across five professional domains (financial, legal, medical, scientific, DevOps) built on real retrieved documents -- actual SEC filings, Federal Register rules, PubMed abstracts, arXiv papers, and GitHub postmortems -- paired with LLM-generated tasks and camouflaged payloads that were not human-validated. Paraphrasing, the strongest defense on synthetic benchmarks, shows no statistically significant attack success rate reduction on real documents (p=0.500) while degrading utility from 91.8% to 82.8%. We introduce PARSE (Provenance-Aware Retrieval Sanitization), a domain-aware, fact-preserving sanitization pipeline that classifies each sentence by injection likelihood, extracts structured facts before rewriting, and verifies fact preservation via a consistency-checking loop. A directiveness gate routes 59% of real enterprise documents to a lightweight path, concentrating computational cost on high-risk documents. PARSE achieves 15.6% attack success rate -- a 39% reduction versus the 25.4% baseline -- at 86.9% utility, the largest reduction of any condition evaluated (h=-0.245), nominally significant (p=0.014) but not surviving correction for multiple comparisons. Practitioners should evaluate defenses on domain-matched real documents, not synthetic proxies.
