DuplexJail: Spoken Interruption Attacks on Full-Duplex Speech Models
cs.CR, cs.SD
Submitted: 2026-09-08
Updated: 2026-09-28
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Moshi: a speech-text foundation model for real-time dialogue
- FLM-Audio: Natural Monologues Improves Native Full-Duplex Chatbots via Dual Training
- BayLing-Duplex: Native Full-Duplex Speech Dialogue with a Single Autoregressive LLM
- Multilingual and Multi-Accent Jailbreaking of Audio LLMs
- Investigating Safety Vulnerabilities of Large Audio-Language Models Under Speaker Emotional Variations
- Universal and Transferable Adversarial Attacks on Aligned Language Models
- HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
- FLEXI: Benchmarking Full-duplex Human-LLM Speech Interaction
- EchoChain: A Full-Duplex Benchmark for State-Update Reasoning Under Interruptions
- Omni-SafetyBench: A Benchmark for Safety Evaluation of Audio-Visual Large Language Models
- RedVox: Safety and Fairness Gaps in Speech Models Across Languages
Related papers
- SoK: AI-Augmented Binary Reversing
- Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
- Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
- Efficient Fuzzy PSI under One-Sided Assumptions
- Sealing the Audit-Runtime Gap for LLM Skills
- Token Composition: A Graph Based on EVM Logs