You’re Now in Control of the Algorithm: How Spotify’s AI Prompts Are Changing Music Discovery

Spotify is experimenting with a new feature that could change how listeners interact with music recommendations, giving users a more direct role in shaping the platform’s algorithm.

‍

The tool, currently being tested with Premium users in select markets, allows listeners to create playlists using simple text prompts. Instead of relying solely on predefined moods or genres, users can describe what they want to hear in their own words. Prompts can be broad, emotional, situational, or highly specific, and Spotify’s system builds a playlist based on both the request and the user’s full listening history.

‍

This approach marks a shift away from passive discovery. Rather than accepting whatever the algorithm serves up, listeners can actively guide it, refining prompts, adjusting the tone, or regenerating playlists until the result feels right. Spotify also plans to let users save prompts and refresh playlists over time, allowing collections to evolve without starting from scratch.

‍

Another notable aspect of the feature is transparency. Spotify shows why certain tracks are included, helping users understand how their prompt and listening behavior influenced the results. This added context aims to make recommendations feel less mysterious and more collaborative.

‍

From Spotify’s perspective, the feature reflects a broader belief that personalization works best when users have agency. Instead of trying to predict what listeners want in the background, the platform is inviting them to participate directly in the recommendation process.

‍

If widely adopted, this kind of prompt-based interaction could signal a larger shift in how streaming platforms handle discovery — moving closer to a co-creation model where human intent and machine learning work together, rather than one replacing the other.

‍

Source: Music Business Worldwide

‍

‍