Spotify: software-implemented inventions and their limit with data protection

March 2021

It is a reality that the SPOTIFY application is part of our lives when we need a moment of disconnection or even while we work. Background music is always a good ally. But have you ever wondered what is behind this application in terms of industrial and intellectual property? It is clear that for this tool to work as it does there is a great work of engineers and computer scientists, among other professionals, behind it.

Spotify has so many functions that we could dedicate an article to each of them, but in general terms, Spotify is a music, podcast and digital video streaming service that gives us access to millions of songs and other content from artists around the world.

In a 250-page report filed with the Securities and Exchange Commission (SEC), the U.S. stock market regulator, the owners explained how the application works, which basically consists of offering all the services mentioned above thanks to the computer storage capacity supported by its gigantic database, which provides information on content consumption and user behavior. To get an idea, this dataset exceeds 200 petabytes, while Netflix, for example, is around 60 petabytes[1].

What they do is track user behaviors such as, for example: song playback, what music we share, how much recommended music we select, which one we skip, which files we actively participate in via vote buttons, etc. Therefore, Spotify’s algorithm anticipates our preferences by using factors such as demographics, past listening behavior and even situational context such as time of day and/or location.

However, for all this to take place, the existence of various computer programs or software is necessary and, in IP matters, there is no little controversy as regards their protection.

LEGAL PROTECTION OF A COMPUTER PROGRAM OR SOFTWARE: PATENT OR COPYRIGHT?

Although throughout history it has become clear that patent law is obviously the most effective framework for protecting an invention, there are several countries in which a distinction is made between inventions embodied in physical media and software-implemented inventions, the latter being protected by copyright. In this sense, although the advantage of copyright is that protection is automatic and free of charge, only the source code or object code is protected against literal copying, but the underlying invention created by the computer program is not protected.

In Europe, although Article 52(2)(c) and (3) excludes from patentability, in principle, a computer program claimed “as such”, after a detailed analysis of the relevant articles, the Board finally concluded that some computer programs could indeed be patented without violating Article 27 of the TRIPS Agreement. This is because those computer programs “as such” only refer to programs that are not of a technical nature. However, as we all know, in order for an invention to be registered and granted its corresponding patent, it must have technical character, be new and make an inventive technical contribution to the knowledge available on the date on which the first application was filed (= priority date), therefore, as long as a computer program has technical character, it is irrelevant the medium in which it is registered, being consequently patentable. While, on the other hand, in the USA, patent protection for inventions containing software is limited to those that are recorded on recordable media and not to computer programs as such[2].

Japan, unlike Europe, the vast majority of Latin America and the USA, has a different patent law that makes specific reference to computer programs as patentable subject matter, which is perhaps the reason why, given the imminent digitalization movement, it is one of the countries with the most patent applications filed in the last year.

Therefore, as we said, the protection granted by a patent is much greater than that granted by copyright, and that is why Spotify has always opted for the first option. Only in 2020 it has filed no less than 56 patent applications. Although only two of them have been granted by the moment.

MUSIC SUGGESTION BASED ON THE USER’S MOOD THROUGH VOICE RECOGNITION – HOW IS MY DATA PROTECTED?

Last February 2021, it was granted to Spotify a patent filed in 2018, which would consist in that Spotify could suggest music based on the recognition of our voice, according to aspects such as the tone of your voice, stress, or the rhythm of speech, to determine, through them, the emotional state, gender, age and even the accent of the user, which leads us at first to think about privacy and obtaining user data that, at the very least, is questionable.

However, if we go to Spotify’s Terms of Use and Privacy Policy – that one that we all with a click claim to have read -, among the data that said company informs us that it will collect, it expressly refers to “Voice Data”, going even more in detail in its “Voice Control Policy” in which it states that:

Spotify can be controlled by your voice. By activating voice controls, you can enable various features, such as searching for songs and playlists with your voice and even interacting with some ads. When you use voice features, your voice input and other information will be sent to Spotify. This includes audio recordings and transcripts of what you say, and other related information, such as the content Spotify showed you. (…). Spotify uses this data to recognize what you say, understand it better and help us provide a response or perform an action, use your data to improve voice functionality on Spotify, develop new voice features and help us provide you with advertising that is more relevant to you. It also includes sharing information from time to time with our service providers, such as cloud storage providers, in order to provide voice functionality to Spotify.”

Notwithstanding the above, it also tells us that:

Spotify will start receiving your voice data when you press the button to activate that feature or say the activation word, or (in the case of voice announcements) when you hear an audible tone and continue until Spotify has processed your question or request.

The device will always indicate when Spotify is receiving your voice input, e.g. with a visual indicator or an audible tone.

You can disable Spotify’s access to the microphone on the device at any time via the operating system settings on the device or by muting the microphone on the device. You can also disable voice announcements at any time in the Spotify settings.”

It seems that, at least for the moment, it gives us the alternative to disable the option for the company to collect our voice data, since, as we know, and after the approval of Regulation (EU) 2016/679 on Data Protection, the user’s consent for the processing of their data has to be express and unambiguous, but to what extent would Spotify’s new patent be functional if all users disable the collection of Voice Data?

Although we do not know if Spotify will finally implement this technology in its service, since the fact that the patent has been granted to the music giant does not imply that it will be implemented (not all inventions patented by large companies usually eventually get implemented; especially considering the hundreds of patent applications that this company has filed), once again the privacy of our data is in question.

What started with a: to register we need your “Name and surname” and “Email”, now it goes as far as the recognition of our voice, our mood, the exact time and place where we are at any given moment. What will be the next barrier to be broken?


[1] 2016 and 2017 data that may have changed as of today’s date.

[2] WIPO Magazine: The patent protection of computer patented inventions, Ania Jedrusik and Phil Wadsworth, 2017.