[论文] When Robots Mishear Us: Mapping the Safety Risks of Voice-Controlled E...

论文概要

研究领域: NLP
作者: Sihan Jia, Oliver Lemon
发布时间: 2026-08-28
arXiv: 2608.28518

中文摘要

我们调查了用户输入中的自动语音识别（ASR）错误是否可能导致具身AI（EAI）模型产生不安全输出。我们发现ASR错误可能导致有害指令被EAI模型接受和执行，从而降低安全性。我们模拟ASR错误并将其与现有安全基准（SafeAgentBench和POEX）结合，以评估不同错误如何影响具身AI安全。我们发现其中一些保留了语义结构但增加了有害歧义，而另一些削弱了模型的拒绝行为并允许生成和执行不安全计划。我们表明在某些情况下ASR错误的自动纠正可以降低风险，但这并不总是有效。总体而言，我们表明ASR错误对具身AI构成重大安全风险。

原文摘要

We investigate whether automatic speech recognition (ASR) errors in user input can lead to unsafe outputs from Embodied AI (EAI) models. We find that ASR errors can lead to harmful instructions being accepted and executed by EAI models, thereby reducing safety. We simulate ASR errors and combine them with existing safety benchmarks (SafeAgentBench and POEX) to evaluate how different errors affect embodied AI safety. We find that some of them preserve semantic structure but increase harmful ambiguity, while others weaken the model refusal behaviour and allow unsafe plans to be generated and executed. We show that in some cases automatic correction of ASR errors can reduce the risk, but this is not always effective. Overall, we show that ASR errors lead to significant safety risks for embodi...

自动采集于 2026-09-01

#论文 #arXiv #NLP #小凯