arXiv Computation and Language

When Helpful Context Leaks: Privacy Risks in Domain-Adapted ASR

SpeechLLMs used in professional settings often undergo domain customisation through prompts or fine‑tuning, which can inadvertently cause the model to transcribe phonetically similar words from its context or training data, leaking private information. The authors systematically investigate this overlooked privacy risk, creating benchmarks to measure leakage rates for both prompting and fine‑tuning, and find that both mechanisms cause measurable leakage that compounds when combined. They evaluate a prompt‑level mitigation strategy and analyse the accuracy‑leakage trade‑off, concluding that fine‑tuning without context prompts offers the best balance between performance and privacy.

arXiv AI
Sep 4

VoxPrivacy: A Benchmark for Evaluating Interactional Privacy of Speech Language Models

The paper introduces VoxPrivacy, a benchmark for assessing interactional privacy in Speech Language Models (SLMs). It evaluates models on a 32‑hour bilingual dataset across three difficulty tiers, revealing that most open‑source SLMs perform near random on conditional privacy decisions and even strong closed‑source systems struggle with proactive privacy inference. The authors also validate these findings on a real‑speech subset and show that fine‑tuning on a 4,000‑hour training set can improve privacy‑preserving capabilities while maintaining robustness.

By Yuxiang Wang, Hongyu Liu, Dekun Chen, Xueyao Zhang, Zhizheng Wu
arXiv Machine Learning
Sep 3

Hearing the Whispers: Black-Box Membership Inference Attacks on Finetuned TTS Models

The paper introduces a black-box membership inference attack framework tailored for fine-tuned text-to-speech models, addressing challenges in query generation and representation engineering. It evaluates five query types, finding recitation queries most effective, and uses multi-level speech embeddings with temporal alignment for fine-grained comparison. Experiments on CosyVoice2, F5-TTS, and XTTS-v2 trained on VCTK and British Dialect datasets show high privacy leakage, with speaker-level AUC above 0.80 and record-level AUC between 0.80 and 0.90.

By Kunlin Cai, Kaiyuan Zhang, Zihang Xiang, Jinghuai Zhang, Abeer Alwan, Fnu Suya, Yuan Tian
arXiv Machine Learning
Aug 27

Reconstruction of Personally Identifiable Information from Proprietary Data in Supervised Fine-Tuned Models

The paper examines how supervised fine-tuning (SFT) of large language models can leak personally identifiable information (PII) when the fine-tuning data contains user-provided sensitive details. It introduces COVA, a coverage-aware decoding algorithm that improves targeted PII reconstruction from SFT models, especially when an adversary has limited contextual knowledge about a target. Experiments on medical and legal Q&A datasets show that even small proprietary SFT datasets can lead to significant privacy leakage via PII reconstruction.

By Sae Furukawa, Alina Oprea
arXiv AI
Sep 4

Anonymization, Not Elimination: Utility-Preserved Speech Anonymization

The paper introduces a two‑stage speech anonymization framework that preserves both linguistic content and acoustic identity. It replaces personally identifiable information using a generative editing model and applies a flow‑matching anonymization technique (F3‑VA) to create diverse, distinct anonymized speakers. The authors evaluate privacy with speaker verification metrics and utility by training ASR, TTS, and SER models from scratch, showing stronger privacy protection with minimal utility loss compared to existing baselines.

By Yunchong Xiao, Yuxiang Zhao, Ziyang Ma, Shuai Wang, Kai Yu, Jiachun Liao, Xie Chen
arXiv Computation and Language
Sep 14

I Am No One: Style-Aware Paraphrasing for Text Anonymization

The paper introduces a style-aware paraphrasing method for text anonymization that leverages pretrained large language models to build compact stylistic profiles from minimal samples and rewrite text to suppress identifiable style markers while preserving meaning. It demonstrates that this approach reduces authorship attribution F1 scores by 60‑70% on blog and review datasets, outperforming both differential privacy‑based and non‑DP baselines, and maintains content quality and readability.

By Ahmed Sohair Khan, Estrid He, Monica Wachowicz, Elham Naghizade
arXiv AI
Aug 19

The Model's Tell: Measuring Context-Leakage Attack Signals with Behavior Gauges

The paper introduces LeakGauge, a method that appends a suffix to a model’s input to gauge the risk of context leakage before decoding. By mapping prefill token probabilities to an attack‑risk score, LeakGauge achieves high AUROC (0.944–0.996) across 11 large language models, including GLM‑5.2 and Kimi‑K3, and remains robust to language changes and different attack styles. The approach also demonstrates sensitivity to internal leakage directions and can be implemented with fewer than 0.5K additional parameters and minimal latency.

By Maosen Zhang, Jianshuo Dong, Boting Lu, Wenyue Li, Xiaoping Zhang, Tianwei Zhang, Jie Zhang, Han Qiu
Hugging Face Trending Papers
Aug 20

Inadvertent Context Leakage in Language Models

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.

arXiv Computation and Language
Aug 28

Your Voice Cloning System is Secretly a Voice Anonymizer

The paper demonstrates that the multilingual voice cloning model XTTSv2 can be repurposed for speaker anonymization without retraining. By conditioning on a pseudo-speaker and using an iterative refinement strategy, the authors balance privacy and intelligibility, achieving near‑optimal privacy (EER ≈ 0.49) and competitive speech quality across seven European languages. The method outperforms dedicated anonymization baselines and requires no language‑specific training.

By Romolo Muletta, Felix Matthias Saaro, Mark Cieliebak, Jan Deriu