[2604.08867] AudioGuard: Toward Comprehensive Audio Safety Protection Across Diverse Threat Models

Computer Science > Sound arXiv:2604.08867 (cs) [Submitted on 10 Apr 2026] Title:AudioGuard: Toward Comprehensive Audio Safety Protection Across Diverse Threat Models Authors:Mintong Kang, Chen Fang, Bo Li View a PDF of the paper titled AudioGuard: Toward Comprehensive Audio Safety Protection Across Diverse Threat Models, by Mintong Kang and 2 other authors View PDF HTML (experimental) Abstract:Audio has rapidly become a primary interface for foundation models, powering real-time voice assistants. Ensuring safety in audio systems is inherently more complex than just "unsafe text spoken aloud": real-world risks can hinge on audio-native harmful sound events, speaker attributes (e.g., child voice), impersonation/voice-cloning misuse, and voice-content compositional harms, such as child voice plus sexual content. The nature of audio makes it challenging to develop comprehensive benchmarks or guardrails against this unique risk landscape. To close this gap, we conduct large-scale red teaming on audio systems, systematically uncover vulnerabilities in audio, and develop a comprehensive, policy-grounded audio risk taxonomy and AudioSafetyBench, the first policy-based audio safety benchmark across diverse threat models. AudioSafetyBench supports diverse languages, suspicious voices (e.g., celebrity/impersonation and child voice), risky voice-content combinations, and non-speech sound events. To defend against these threats, we propose AudioGuard, a unified guardrail consisting of ...