Whenever we talk about smart devices, the question arises whether they could compromise our privacy. After all, they are devices that are permanently connected to the network. The truth is that yes, they could be a problem, especially if we do not have them properly configured. In this article, we are going to talk about the specific case of smart speakers. Where does everything we say go? Could they be a problem?
Smart speakers are quite popular when we start home automation. They are easy to use, useful and there are different models. But of course, they are a perfect example of a device that could compromise privacy, if we take into account that they are listening continuously. But is it really like that?
What happens to what you say to a smart speaker
The first thing you should know is that, unless there is a problem, the speakers only start listening once they detect the “magic word.” For example, saying “hey, Siri” or “Alexa”. It will only use everything you say from that moment on. You can store that and remember it in the future. In addition, it is normal for this type of speakers to activate the light when they are listening.
But what happens to what you say once the speaker has been activated? In that case, it does go outside your network. What you say will go to a server hosted in the cloud, where it will be processed to provide the necessary information. For example, if you say something like “Alexa, tell me if it’s going to rain tomorrow.” That will end up on a remote server.
Basically, your voice goes to an external server, where the information is processed and the speaker emits a response. It could be a question you have asked, something you have told him to remember you in the future, etc. There are speakers, as in the case of Apple, that this information, if it goes to a remote server, is previously encrypted.
You must also keep in mind that the information you have sent, your voice, in short, may be kept for a while. This will depend on the speaker, as well as the message. But yes, it may remain for a while in the cloud, on servers owned by that manufacturer.
A human can hear you
But not only should it be mentioned that the speakers send the sound to the cloud or can store it for a while, but a human could also hear you. In this case, it serves to perfect the system, find errors so that these speakers can work better.
It does not mean that they listen to everything you say, but only a very small part of all users. However, this will depend on the speaker. There are some models that will listen without you having to do anything, while in others you will have to give permissions so that they can use that data to improve the system.
Ultimately, what you say to a smart speaker may or may not leave your network. Depending on the model, it may even be stored for a time in the cloud, on servers owned by the manufacturer. A human could also even hear you, in certain cases.
