Ban on using audio features of a song for ML or algorithms

[alex_192]

Can someone shed some light on why and when the audio features from the API were banned from being used for Machine Learning or any algorithms in general? I'm curious because there are a lot of papers that have used them for that, but maybe the ban is recent? Thank you.

[Ximzend · Accepted solution]

Hi @alex_192 ,

The ban you're referring to was outlined in the updated Spotify Developer Policy, which took effect on May 8, 2023. This change was highlighted in the Spotify Developer blog on May 29, 2023.

So, to answer your question: the restriction came into place in May 2023. Prior to that, it was common to see papers and projects utilizing Spotify's audio features for machine learning.

I hope this clarifies things! Let me know if you have any further questions.