Auckland: The power of platforms to suggest content we may like is a boon for convenience, but secrecy around data collection raises troubling questions. Billions of people receive recommendations from digital platforms that impact on their daily offline life. What series they binge next, where they go on holiday, their next favourite song and the people they date are often shaped by the algorithmic curation of digital platforms.
Platforms need to know as much as possible about their users in order to make these recommendations. But the full extent of the data that companies collect and use is still largely a closely guarded secret. Recent research has found that Spotify and Tinder are extracting an ever-growing amount of data from users. Through an analysis of successive versions of the platforms' terms of use and privacy policies, the research shows the growing types of data taken: in Spotify's case, photos, location, voice data, and personal information like payment details are being collected.
Both companies vaguely list the types of user data they collect in Spotify's case, data for marketing, promotion, and advertising purposes. But why would a platform like Spotify use voice data? On its website, the company lists several reasons for collecting various types of data, but these explanations, like many other platforms', are vague and open to interpretation.
But even if the reasons were clearly stated, there is no way of knowing whether what we're being told is true. The underlying code driving the algorithms is protected and platforms are under no legal obligation to share it. In its 2021 policy, Spotify explicitly said that some user data is used in its recommendations. The company reserves the right to make inferences about users' interests, and says the content [users] view is determined not just by their data but by Spotify's commercial agreements with third parties. What Spotify markets as highly personalised, tailored content is influenced by factors outside users' listening data, including agreements with artists and labels.
This two-pronged approach is a pincer movement: an exploitative strategy that simultaneously attacks user privacy and rights while defining their tastes, choices and identity. The first movement is about extracting as much data and information from users as possible. Once this data is processed, the second movement is about conditioning users' online and offline behaviours.