AI RESEARCH
Privacy-Preserving End-to-End Full-Duplex Speech Dialogue Models
arXiv CS.AI
•
ArXi:2603.08179v1 Announce Type: cross End-to-end full-duplex speech models feed user audio through an always-on LLM backbone, yet the speaker privacy implications of their hidden representations remain unexamined. Following the VoicePrivacy 2024 protocol with a lazy-informed attacker, we show that the hidden states of SALM-Duplex and Moshi leak substantial speaker identity across all transformer layers.