Spotify is the world's largest audio streaming service, with over 600 million users. It uses AI extensively for music recommendations, podcast discovery, and now AI-generated DJ experiences.
Spotify's AI DJ feature uses a synthesized voice modeled on real DJ Xavier 'X' Jernigan to deliver personalized music commentary, with the voice cloned using AI voice synthesis technology.
The AI DJ voice was created in partnership with SoundLabs, using voice cloning technology applied to recordings of Spotify's head of cultural partnerships. The DJ feature generates contextual commentary scripts that the synthesized voice reads aloud.
Spotify's recommendation engine uses machine learning to power Discover Weekly, Daily Mixes, and the home feed, analyzing listening history, collaborative filtering across millions of users, and audio feature extraction.
Spotify's system combines three approaches: collaborative filtering (users with similar taste), NLP on playlist names and blog text, and direct audio analysis (tempo, key, timbre) via convolutional neural networks.
Have evidence about Spotify's AI practices? Submit it.
Report a Sighting