Telemetry or cyber spying?
Just for a minute imagine yourself being a developer who’s rolled out a device or application. Of course, now you want to learn what functions users will be using and how often, what issues and errors they will encounter. Such information will allow you to enhance your solution and drive up sales.
To accomplish that task, you have several choices. You can survey your users and collect their answers. In this event a user provides only the data he wants to share. We use this method collecting information about Panic Button. You can take a look at our survey at https://panicbutton.pw/survey/.
However, this method has obvious disadvantages: first, users share their information reluctantly, second, they are often unable to objectively assess the information about the functionality they use. Third, many errors are unseen and unclear to users, and they can be discovered only using software.
The advantages: no unsanctioned user data collection is taking place in this case, and you also have an opportunity to explore user satisfaction telemetry can find out only indirectly.
An automated collection of telemetry data has indisputable advantages over surveys: it provides all the required information about user, including the system’s version, configuration and application’s version, location, consumption of resources, CPU load and a lot of other information. You can get a detailed list of all the options being used, time of use, technical data about all errors and application crashes.
From there all the information is sent to the server and automatically processed. This raises an interesting question, Isn’t the line between telemetry collection and cyber spying too fine? Should you link the information to a certain user or just impersonalize data?
For instance, if the application collects the information about all the sessions you have launched and send it without linking it to you specifically, we are dealing with telemetry. If the data is somehow linked with you, we are talking about cyber spying. The application can know your email, but it is not required for collecting data to improve the application. However, your email is required if it performs a targeted collection of information about you.
In our opinion, telemetry should be collected without any identification of users, be it their email or IP address. It is possible to use an identifier that’s not linked with the user whatsoever, that shouldn’t be static and should be updated in a given amount of time or every time a new session is run.
When the data are delivered and processed, they should be impersonalized... However, you understand that it is impossible to check everything. Anyway, this data doesn’t arrive anonymously, the IP address of the device from where it is sent is linked to this data. This IP address may be further removed, as Mozilla promises us, or saved though it doesn’t offer any value as far as application improvements are concerned. One way or another, you won’t be able to check if the information about your IP address is deleted or not.
Many experts, when distinguishing between cyber spying and telemetry collection, recommend sticking to the goal of data collection and use it as the basis for separating cyber spying and telemetry data collection. If you pursue the goal of enhancing the performance of your application, get an insight into how users use it – you are involved in telemetry. If you target users themselves, their data and activity, it’s cyber spying. In my opinion, this is arguable though I can’t help agreeing to some of this rationale.
For instance, if the browser collects data about all the websites the user has visited, this hardly resembles telemetry as similar information won’t affect the performance of the application at all, this is just data collection that will be probably sold off for profit.
The bad thing is when you encounter closed source software because in this case you won’t know what data exactly the application is sending as they are often held encrypted. Of course, tracking queries will allow you to identify where they are going and how frequently, but this information is obviously not enough.
So we can’t check what exactly the application is sending and how it stores data. As a rule, this is mentioned in the privacy notice, however, as you have probably guessed, it may contain unreliable information.
Examples of telemetry data collection
Let’s distinguish telemetry by the source used for obtaining it: data collection by programs, operating systems and devices. Of course, when it comes to devices, telemetry data is collected by the software pre-installed by the developer.
Let’s start with a program and use the browser Mozilla Firefox as an example. You won’t believe me if I told you what their policy notice says, so let me just copy the information the browser passes on about you by default (available on the official website).
Interaction data: Firefox sends data about your interactions with Firefox to us (such as number of open tabs and windows; number of