Inadvertent Context Leakage in Language Models

• URL: http://arxiv.org/abs/2608.19857v1
• Date: Thu, 20 Aug 2026 10:05:29 GMT
• Title: Inadvertent Context Leakage in Language Models
• Abstract summary: For AI agents to be useful beyond simple chat, they must hold sensitive user context such as calendars, credentials, health records, and financial data.<n>We study whether the mere presence of such secrets in a model's context window introduces hidden correlations into the model's benign outputs.<n>We show that this limited leakage is exploited using a novel adaptive attack that assumes black-box access to the underlying model.
• Score: 44.48836485611056
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: For AI agents to be useful beyond simple chat, they must hold sensitive user context such as calendars, credentials, health records, and financial data. We study whether the mere presence of such secrets in a model's context window introduces hidden correlations into the model's benign outputs, allowing reconstruction even when the model correctly refuses direct extraction. We further study whether an adversary can actively engineer prompts that amplify this effect, using the model as a covert carrier to transmit secrets through seemingly innocuous text. In both cases, this limited leakage is exploited using a novel adaptive attack that assumes black-box access to the underlying model. In controlled experiments across eight proprietary models, we find that 2-digit in-context secrets are reconstructed with near-perfect accuracy and 4-digit secrets at 82\% exact match, all from outputs the model produces in response to ordinary, non-adversarial requests. We observe that more capable models leak more: stronger instruction-following amplifies sensitivity to in-context secrets, suggesting leakage is a byproduct of capability as opposed to a patchable bug. We show this leakage enables two practical attacks: (1) a trained classifier that infers semantic predicates about user memories (e.g., health conditions, financial events) from routine natural-language outputs, and (2) an RL-trained adversary that extracts full Social Security Numbers from a production-style agent.

Related papers

• Evaluating Language Model Reasoning about Confidential Information [95.64687778185703]
  We study whether language models exhibit contextual robustness, or the capability to adhere to context-dependent safety specifications.<n>We develop a benchmark (PasswordEval) that measures whether language models can correctly determine when a user request is authorized.<n>We find that current open- and closed-source models struggle with this seemingly simple task, and that, perhaps surprisingly, reasoning capabilities do not generally improve performance.
  arXiv  Detail & Related papers  (2025-08-27T15:39:46Z)
• Watch the Weights: Unsupervised monitoring and control of fine-tuned LLMs [27.544312683007234]
  We introduce a new method for understanding, monitoring and controlling fine-tuned large language models (LLMs)<n>We demonstrate that the top singular of the weight difference between a fine-tuned model and its base model correspond to newly acquired behaviors.<n>For backdoored models that bypasses safety mechanisms when a secret trigger is present, our method stops up to 100% of attacks with a false positive rate below 1.2%.
  arXiv  Detail & Related papers  (2025-07-31T21:04:12Z)
• ExpShield: Safeguarding Web Text from Unauthorized Crawling and Language Modeling Exploitation [17.71790411163849]
  We propose ExpShiled, a proactive self-defense mechanism that mitigates sample-specific memorization via imperceptible text perturbations.<n>Our approach requires no external collaboration while maintaining original readability.<n>Even with privacy backdoors, the Membership Inference Attack (MIA) AUC drops from 0.95 to 0.55, and instance exploitation approaches zero.
  arXiv  Detail & Related papers  (2024-12-30T17:52:02Z)
• Navigating the OverKill in Large Language Models [84.62340510027042]
  We investigate the factors for overkill by exploring how models handle and determine the safety of queries. Our findings reveal the presence of shortcuts within models, leading to an over-attention of harmful words like 'kill' and prompts emphasizing safety will exacerbate overkill. We introduce Self-Contrastive Decoding (Self-CD), a training-free and model-agnostic strategy, to alleviate this phenomenon.
  arXiv  Detail & Related papers  (2024-01-31T07:26:47Z)
• Scalable Extraction of Training Data from (Production) Language Models [93.7746567808049]
  This paper studies extractable memorization: training data that an adversary can efficiently extract by querying a machine learning model without prior knowledge of the training dataset. We show an adversary can extract gigabytes of training data from open-source language models like Pythia or GPT-Neo, semi-open models like LLaMA or Falcon, and closed models like ChatGPT.
  arXiv  Detail & Related papers  (2023-11-28T18:47:03Z)
• Leveraging Diffusion-Based Image Variations for Robust Training on Poisoned Data [26.551317580666353]
  Backdoor attacks pose a serious security threat for training neural networks. We propose a novel approach that enables model training on potentially poisoned datasets by utilizing the power of recent diffusion models.
  arXiv  Detail & Related papers  (2023-10-10T07:25:06Z)
• ParaFuzz: An Interpretability-Driven Technique for Detecting Poisoned Samples in NLP [29.375957205348115]
  We propose an innovative test-time poisoned sample detection framework that hinges on the interpretability of model predictions. We employ ChatGPT, a state-of-the-art large language model, as our paraphraser and formulate the trigger-removal task as a prompt engineering problem.
  arXiv  Detail & Related papers  (2023-08-04T03:48:28Z)
• Large Language Models with Controllable Working Memory [64.71038763708161]
  Large language models (LLMs) have led to a series of breakthroughs in natural language processing (NLP) What further sets these models apart is the massive amounts of world knowledge they internalize during pretraining. How the model's world knowledge interacts with the factual information presented in the context remains under explored.
  arXiv  Detail & Related papers  (2022-11-09T18:58:29Z)
• MOVE: Effective and Harmless Ownership Verification via Embedded External Features [104.97541464349581]
  We propose an effective and harmless model ownership verification (MOVE) to defend against different types of model stealing simultaneously.<n>We conduct the ownership verification by verifying whether a suspicious model contains the knowledge of defender-specified external features.<n>We then train a meta-classifier to determine whether a model is stolen from the victim.
  arXiv  Detail & Related papers  (2022-08-04T02:22:29Z)
• Combing for Credentials: Active Pattern Extraction from Smart Reply [15.097010165958027]
  We investigate potential information leakage vulnerabilities in a typical Smart Reply pipeline. We introduce a new type of active extraction attack that exploits canonical patterns in text containing sensitive data. We show experimentally that it is possible for an adversary to extract sensitive user information present in the training data, even in realistic settings.
  arXiv  Detail & Related papers  (2022-07-14T05:03:56Z)

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.