Privacy Leakage in Language Models

Overview

The field studies when language models reveal information they learned or received, including personal details, training-set membership, and sensitive traits inferred from text. Researchers test verbatim extraction and membership inference, as well as reconstruction from gradients, activations, hidden states, and model outputs; retrieval systems, persistent memory, and tool-using agents create additional risks across turns and users. Defenses include differential privacy, data filtering, unlearning, activation steering, prompt sanitization, obfuscation, and system-level access controls and information-flow enforcement. A central tension is that protections can reduce leakage while harming usefulness, and refusal or successful output-level forgetting does not necessarily prevent information from being recovered through other prompts, representations, or system channels.

Latest papers

303 papers22 papers in the last four weeks, down 54% on the four weeks before. 0.2% of all new papers.

Oct 3, 2024

Mitigating Memorization In Language Models

Language models (LMs) can "memorize" information, i.e., encode training data in their weights in such a way that inference-time queries can lead to verbatim regurgitation of that data. This ability to extract training data can be problematic, for example, when data are private or sensitive. In this work, we investigate methods to mitigate memorization: three regularizer-based, three finetuning-based, and eleven machine unlearning-based methods, with five of the latter being new methods that we introduce. We also introduce TinyMem, a suite of small, computationally-efficient LMs for the rapid development and evaluation of memorization-mitigation methods. We demonstrate that the mitigation methods that we develop using TinyMem can successfully be applied to production-grade LMs, and we determine via experiment that: regularizer-based mitigation methods are slow and ineffective at curbing memorization; fine-tuning-based methods are effective at curbing memorization, but overly expensive, especially for retaining higher accuracies; and unlearning-based methods are faster and more effective, allowing for the precise localization and removal of memorized information from LM weights prior to inference. We show, in particular, that our proposed unlearning method BalancedSubnet outperforms other mitigation methods at removing memorized information while preserving performance on target tasks.
Jun 8, 2024

Could Mobile AI be Exploited?: Understanding the Exploitations of Trust and Ethics Gap in On-Device Small Language Models

In this paper, we present a very first study to investigate trust and ethical implications of on-device artificial intelligence (AI), focusing on small language models (SLMs) amenable for personal devices like smartphones. While on-device SLMs promise enhanced privacy, reduced latency, and improved user experience compared to cloud-based services, we posit that they might also introduce significant risks and vulnerabilities compared to their on-server counterparts. As part of our study, we conduct a systematic evaluation of the state-of-the-art on-devices SLMs, compared to their on-server counterparts, using well-established trust and ethics frameworks. Our results illustrate that on-device SLMs are significantly less trustworthy and lacks ethical safeguards, specifically demonstrating more stereotypical, unfair, privacy-breaching and harmful content generation behavior. Further, these risks of on-device SLMs could be exploited through Open Knowledge (OkNow) Attacks, and we demonstrated them using potentially unethical vanilla prompts, to which the on-device SLMs answer with valid responses without any filters and without the need for any jailbreaking or prompt engineering. These responses can be abused for various harmful and unethical scenarios like: societal harm, illegal activities, hate, self-harm, exploitable phishing content, disinformation and many others.
Mar 10, 2024

PPFedIT: Towards Privacy-Preserving Federated Instruction Tuning with Few-shot Local Examples

Instruction tuning aligns large language models (LLMs) with human intentions but requires diverse, high-quality data that are difficult to collect in privacy-sensitive domains. Federated instruction tuning (FedIT) enables collaborative training across data owners, yet existing methods typically assume sufficient local data. In realistic few-shot settings, limited samples can cause overfitting, degrade performance, and increase vulnerability to training data extraction attacks. We propose PPFedIT, a federated algorithm that improves both model performance and privacy protection in federated few-shot learning. It comprises three client-side steps: (1) synthetic data generation, which uses LLMs to diversify and enrich local data; (2) parameter isolation training, which updates the shared global LLM on synthetic data and local LLMs on private local data to mitigate synthetic-data noise; and (3) local aggregation then sharing, which mixes global and local model parameters before uploading them for server aggregation to mitigate data extraction attacks. Experiments on three open-source datasets show that PPFedIT improves model performance by an average of 8.4% and reduces the risk of data extraction attacks by approximately 20% in challenging federated few-shot settings.