Spotify had to explain how its voice assistant uses users audio data. This happened almost immediately after the function was started. Users recalled the data published at the beginning of the year about the Spotify patent for algorithms for determining the age, gender, accent, and even the emotional state of the speaker by voice. All this is hardly necessary for recognizing commands about choosing music.
Indeed, in the Spotify voice assistant FAQ, you can find information that the information obtained can be used for “more relevant”, targeted, that is, targeted advertising. This is unlikely to affect Premium subscribers, but even they should pay attention to one subtlety: Spotify acknowledged that the firm will share data with partners “when it is necessary for Spotify to work.”
However, the streaming service guarantees that the microphone signal is written only when the wake-up command is received-with a button or the phrase “Hey Spotify”. After” waking up”, the voice control waits for a few seconds and, if no commands are heard during this time, it goes back to standby mode. According to Spotify, while the system is waiting for the “Hey Spotify” command, it does not record anything.