# VoiceSHIELD-Small: Real-Time Malicious Speech Detection and Transcription

Source: [VoiceSHIELD-Small: Real-Time Malicious Speech Detection and Transcription](<https://www.sota2.com/research/paper/69afa58c2d603e6de3e43ad9>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-11T05:00:11.798Z

Authors: Sumit Ranjan, Sugandha Sharma, Ubaid Abbas, Puneeth N Ail

Published: 2026-03-08T16:23:42.000Z

## Abstract

Voice interfaces are quickly becoming a common way for people to interact with AI systems. This also brings new security risks, such as prompt injection, social engineering, and harmful voice commands. Traditional security methods rely on converting speech to text and then filtering that text, which introduces delays and can ignore important audio cues. This paper introduces VoiceSHIELD-Small, a lightweight model that works in real time. It can transcribe speech and detect whether it is safe or harmful, all in one step. Built on OpenAI's Whisper-small encoder, VoiceSHIELD adds a mean-pooling layer and a simple classification head. It takes just 90-120 milliseconds to classify audio on mid-tier GPUs, while transcription happens at the same time. Tested on a balanced set of 947 audio clips, the model achieved 99.16 percent accuracy and an F1 score of 0.9865. At the default setting, it missed 2.33 percent of harmful inputs. Cross-validation showed consistent performance (F1 standard deviation = 0.0026). The paper also covers the model's design, training data, performance trade-offs, and responsible use guidelines. VoiceSHIELD is released under the MIT license to encourage further research and adoption in voice AI security.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Malicious Speech Classification on Internal Speech Safety Dataset (test)](<https://www.sota2.com/research/sota/malicious-speech-classification-on-internal-speech-safety-dataset-test.md>): Malicious Speech Classification; Internal Speech Safety Dataset (test); Accuracy: 99.16.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.07708>)

[PDF](<https://arxiv.org/pdf/2603.07708.pdf>)
