PRIVACY IMPLICATIONS OF SMART SPEAKER INADVERTENT ACTIVATIONS: AN EMPIRICAL STUDY
Authors/Creators
Description
The widespread adoption of smart speakers by leading manufacturers such as Apple, Google, and Amazon highlights the growing usage of voice-activated devices in homes. These devices are incredibly convenient, but they also present significant privacy hazards, particularly regarding unintentional activations and undesired data transmission. Investigating the situations in which smart speakers are inadvertently turned on and evaluating users’ potential privacy risks were the goals of this study. The authors tracked device activations and network data transmission trends by playing 120 hours of video content and utilizing Wireshark to identify undesired data spikes during false activations. The results demonstrated frequent false activations associated with specific words or phrases and associated data transmission activities, suggesting significant privacy concerns. False activation rates varied among Apple, Google, and Amazon speakers, according to device comparisons. This could be a sign of differences in speech recognition accuracy and privacy issues. This study draws attention to the limitations of voice-recognition systems and advocates for technological advancements and regulatory changes to enhance user control and data security when utilizing smart speakers. The authors encourage future researchers to investigate innovative ways to resolve these privacy concerns, ensure greater transparency, and boost user trust in voice-activated devices.
Files
Files
(369.8 kB)
| Name | Size | Download all |
|---|---|---|
|
md5:95bacb68cc4e6d0992c9d2b789b3223b
|
369.8 kB | Download |