One Size Fits All? A Modular Adaptive Sanitization Kit (MASK) for Customizable Privacy-Preserving Phone Scam Detection

2025-10-21

AI Summary

To address privacy leakage risks arising from large language model (LLM) deployment in telephone fraud detection, this paper proposes MASK—a trainable, modular privacy-preserving framework. MASK adopts a plug-in architecture integrating differentiable keyword filtering and neural de-identification modules, enabling dynamic adjustment of privacy protection strength according to user preferences. A configurable loss function jointly optimizes fraud detection accuracy and privacy guarantees. Experiments demonstrate that MASK maintains high fraud detection accuracy while significantly reducing personally identifiable information (PII) exposure risk, and seamlessly interoperates with both rule-based engines and deep learning methods. This work presents the first unified framework for LLM-driven fraud detection that simultaneously achieves programmable privacy strength, differentiable de-identification, and extensible system architecture—providing an efficient and trustworthy solution for privacy-sensitive AI applications.

Technology Category

• Machine Learning: Privacy
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Security and Privacy: Security and privacy of machine learning and AI applications

Abstract

Phone scams remain a pervasive threat to both personal safety and financial security worldwide. Recent advances in large language models (LLMs) have demonstrated strong potential in detecting fraudulent behavior by analyzing transcribed phone conversations. However, these capabilities introduce notable privacy risks, as such conversations frequently contain sensitive personal information that may be exposed to third-party service providers during processing. In this work, we explore how to harness LLMs for phone scam detection while preserving user privacy. We propose MASK (Modular Adaptive Sanitization Kit), a trainable and extensible framework that enables dynamic privacy adjustment based on individual preferences. MASK provides a pluggable architecture that accommodates diverse sanitization methods - from traditional keyword-based techniques for high-privacy users to sophisticated neural approaches for those prioritizing accuracy. We also discuss potential modeling approaches and loss function designs for future development, enabling the creation of truly personalized, privacy-aware LLM-based detection systems that balance user trust and detection effectiveness, even beyond phone scam context.

Problem

• Detecting phone scams using LLMs while preserving user privacy
• Balancing privacy protection with scam detection accuracy
• Providing customizable privacy preferences for sensitive conversation data

Innovation

• Modular framework enables dynamic privacy adjustment
• Pluggable architecture supports diverse sanitization methods
• Trainable system balances detection accuracy with privacy