A few considerations

The lack of representation of women in the recommendations of both Deezer and Spotify highlights the need for further analysis on how music platforms can contribute to imbalances in relation to historically marginalized groups. Although both services have been mentioned here only as examples of what can be observed when we analyze this kind of consumption, as previously mentioned, the problems related to these inequalities are not limited to flaws or biases present in music platforms; they also draw our attention to these services’ social responsibility, which could contribute to the pursuit of equity in the music industry.

Over the course of the research, in addition to noticing the little representation of women, non-binary people and transvestites on music recommendations, we also came across uncertainties about whether we were facing a fully automated system or if people working at Spotify or Deezer, for example, had directly curated the playlists. The lack of this knowledge underscores once again the importance of making data available, reducing the opacity of information for researchers. Access to this data would be beneficial for the dialogue not only with policymakers, but also with civil society and music platforms.

In addition, the lack of data on black and non-white people in general aggravates the lack of transparency of the platforms. Not having access to how people self-identify, both as users and artists, prevents us from identifying whether streaming platforms reflect the inequalities we observe outside platforms. According to a survey conducted by the journalist Léo Feijó, “in 62.5% of the organizations analyzed, less than 5% of the executive positions are held by black people. Besides, in more than 46% of the organizations in Brazilian music industry, the presence of blacks varies from 0% to 15% of the total number of employees. As for the positions in the total staff, only 26.5% are occupied by blacks.1” Without having this data, how can we know if the scenario of streaming platforms resembles or differs from these statistics?

Although part of the debates about platform regulation, artificial intelligence, hate speech and privacy safety has focused on other aspects of possible negative impacts on society, in general, and on marginalized groups, in particular, we believe it is important to also focus on the influence on the formation of musical taste and on the interrelation of streaming services and users, which can connect to copyright and, to some extent, go beyond it, since the discussion about intellectual property itself is not tangible to a significant part of musicians and of society in general2.  

Music is conceived as a considerable part of the formation and identification of peoples, but it is also directly related to the way artists and other music professionals place themselves in this scenario. For this reason, it is essential to carry out further research on:

1. To which other data, besides race and ethnicity, do we need to have access? How can we balance access to these data with care regarding the privacy of users? 
2. What other social markers of difference are operationalized and come across algorithmic biases? Can sexuality, territoriality and nationality be some of them? 
3. What comparisons can be drawn between the Global North and the Global South? Are there significant differences with respect to data access and the biases found? 
4. What solutions could be found through dialogues among policymakers, music platforms and civil society? 
5. What space can artists, music professionals and users have in future developments and improvements of music apps?