A Case Study on the Impact of Anonymization Along the RAG Pipeline

• URL: http://arxiv.org/abs/2604.15958v1
• Date: Fri, 17 Apr 2026 11:23:59 GMT
• Title: A Case Study on the Impact of Anonymization Along the RAG Pipeline
• Abstract summary: Using anonymization techniques to remove personally identifiable information (PII) and other sensitive markers in the underlying data represents a practical and sensible course of action for RAG administrators.<n>We systematically and empirically measure the impact of anonymization at two important points along the RAG pipeline: the dataset and generated answer.<n>We show that differences in privacy-utility trade-offs can be observed depending on where anonymization took place, demonstrating the significance of privacy risk mitigation placement in RAG.
• Score: 16.51235052027641
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Despite the considerable promise of Retrieval-Augmented Generation (RAG), many real-world use cases may create privacy concerns, where the purported utility of RAG-enabled insights comes at the risk of exposing private information to either the LLM or the end user requesting the response. As a potential mitigation, using anonymization techniques to remove personally identifiable information (PII) and other sensitive markers in the underlying data represents a practical and sensible course of action for RAG administrators. Despite a wealth of literature on the topic, no works consider the placement of anonymization along the RAG pipeline, i.e., asking the question, where should anonymization happen? In this case study, we systematically and empirically measure the impact of anonymization at two important points along the RAG pipeline: the dataset and generated answer. We show that differences in privacy-utility trade-offs can be observed depending on where anonymization took place, demonstrating the significance of privacy risk mitigation placement in RAG.

Related papers

• Profiling for Pennies: Unveiling the Privacy Iceberg of LLM Agents [52.02752011119451]
  We investigate privacy perception in the real world through the lens of human awareness.<n>We propose the PrivacyIceberg, which categorizes real-world human privacy risks into three tiers.
  arXiv  Detail & Related papers  (2026-05-07T13:21:44Z)
• Not All Entities are Created Equal: A Dynamic Anonymization Framework for Privacy-Preserving Retrieval-Augmented Generation [7.905833478464377]
  We introduce a dynamic anonymization framework named TRIP-RAG.<n>It identifies highly sensitive entities while trading off utility.<n>Our theoretical analysis and experiments indicate that TRIP-RAG can effectively reduce context inference risks.
  arXiv  Detail & Related papers  (2026-03-27T05:03:24Z)
• Neural Gate: Mitigating Privacy Risks in LVLMs via Neuron-Level Gradient Gating [71.55435880263238]
  We introduce Neural Gate, a novel method for mitigating privacy risks through neuron-level model editing.<n>Our method improves a model's privacy safeguards by increasing its rate of refusal for privacy-related questions.
  arXiv  Detail & Related papers  (2026-03-13T03:03:20Z)
• Evaluating the Impact of Data Anonymization on Image Retrieval [2.8165396388850112]
  We study the impact of anonymization on Content-Based Image Retrieval (CBIR) using two public datasets and the internal DOKIQ dataset.<n>Our results reveal a pronounced retrieval bias in favor of models trained on original data, which produce the most similar retrievals after anonymization.
  arXiv  Detail & Related papers  (2026-02-23T09:39:06Z)
• A Systemic Evaluation of Multimodal RAG Privacy [28.429497070260027]
  We analyze the privacy risks inherent in the mRAG pipeline observed through standard model prompting.<n>Our findings highlight the need for privacy-preserving mechanisms and motivate future research on mRAG privacy.
  arXiv  Detail & Related papers  (2026-01-25T01:37:01Z)
• SoK: Privacy Risks and Mitigations in Retrieval-Augmented Generation Systems [53.51921540246166]
  Retrieval-Augmented Generation (RAG) techniques have become widely popular.<n>RAG involves the coupling of Large Language Models (LLMs) with domain-specific knowledge bases.<n>The proliferation of RAG has sparked concerns about data privacy.
  arXiv  Detail & Related papers  (2026-01-07T14:50:41Z)
• Fine-Grained Privacy Extraction from Retrieval-Augmented Generation Systems via Knowledge Asymmetry Exploitation [15.985529058573912]
  Retrieval-augmented generation (RAG) systems enhance large language models (LLMs) by integrating external knowledge bases.<n>Existing privacy attacks on RAG systems can trigger data leakage but often fail to accurately isolate knowledge-base-derived sentences within mixed responses.<n>This paper presents a novel black-box attack framework that exploits knowledge asymmetry between RAG and standard LLMs to achieve fine-grained privacy extraction.
  arXiv  Detail & Related papers  (2025-07-31T03:50:16Z)
• Self-Refining Language Model Anonymizers via Adversarial Distillation [48.280759014096354]
  We introduce SElf-refining Anonymization with Language model (SEAL)<n>SEAL is a novel distillation framework for training small language models (SLMs) to perform effective anonymization without relying on external models at inference time.<n>Experiments on SynthPAI, a dataset of synthetic personal profiles and text comments, demonstrate that SLMs trained with SEAL achieve substantial improvements in anonymization capabilities.
  arXiv  Detail & Related papers  (2025-06-02T08:21:27Z)
• A False Sense of Privacy: Evaluating Textual Data Sanitization Beyond Surface-level Privacy Leakage [77.83757117924995]
  We propose a new framework that evaluates re-identification attacks to quantify individual privacy risks upon data release.<n>Our approach shows that seemingly innocuous auxiliary information can be used to infer sensitive attributes like age or substance use history from sanitized data.
  arXiv  Detail & Related papers  (2025-04-28T01:16:27Z)
• Investigating Vulnerabilities of GPS Trip Data to Trajectory-User Linking Attacks [49.1574468325115]
  We propose a novel attack to reconstruct user identifiers in GPS trip datasets consisting of single trips.<n>We show that the risk of re-identification is significant even when personal identifiers have been removed.<n>Further investigations indicate that users who frequently visit locations that are only visited by a small number of others tend to be more vulnerable to re-identification.
  arXiv  Detail & Related papers  (2025-02-12T08:54:49Z)
• The Good and The Bad: Exploring Privacy Issues in Retrieval-Augmented Generation (RAG) [56.67603627046346]
  Retrieval-augmented generation (RAG) is a powerful technique to facilitate language model with proprietary and private data. In this work, we conduct empirical studies with novel attack methods, which demonstrate the vulnerability of RAG systems on leaking the private retrieval database.
  arXiv  Detail & Related papers  (2024-02-23T18:35:15Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.