Full Description
Voice-controllable devices allow hands-free interactions, offering a modern user experience and helping impaired users. However, the voice channel that enables this interaction is also one of their main weaknesses.
Higher usability often comes with higher security and privacy risks. As Voice-Controllable Devices are no exception to this, a plethora of new attacks have been carried out against them in recent years. In this book, several attacks are analysed against the voice channel available in the literature; it then devises a bespoke kill chain to facilitate the classification of attack steps and future research. Additionally, the work analyses the design of the notorious Alexa vs Alexa attack, which turned Echo Dot devices against themselves in 2022. Finally, the present volume explores measures against all these attacks and raise important questions about their suitability for user needs.
Topics and features:
Discusses timely security and privacy challenges in IoT devices, with an emphasis on voice-controllable devices
Provides a framework for assessing and studying attacks against the voice channel
Addresses practical attacks successfully carried out against real devices, showing that an impactful attack need not be complex
Explores advanced safeguards from a new perspective, questioning whether research is proceeding in the right direction
This powerful volume—containing tips, tricks, and techniques used by ethical hackers to breach the voice channel—serves as the perfect guide for researchers conducting security and privacy research on IoT. It also provides food for thought for graduate students and enthusiasts of these subjects who do not shy away from the mathematics behind the world of machine learning.
Contents
1. Introduction.- 2. Background.- 3. The VOCODES Framework for Attacking VCDS.- 4. The Alexa versus Alexa Attack.- 5. An Assessment of AVA in the Real World.- 6. A Taxonomy of Measures Against Voice-Spoofing Attacks.- 7. An Intelligent Measure Against Self-Activation.- 8. Conclusion.
