Spotify says its best developers haven't written a line of code since December, thanks to AI

Spotify Developers Haven't Written Code Since December

Spotify has reached a striking milestone in AI coding adoption. During its fourth-quarter earnings call, co-CEO Gustav Söderström announced that the company's best developers "have not written a single line of code since December."1

1

Despite this dramatic shift, the streaming giant maintained its rapid pace of new feature launches, shipping more than 50 new features and changes throughout 2025.2

2

Recent additions include AI-powered Prompted Playlists, Page Match for audiobooks, and About This Song, all rolled out within recent weeks.

Internal AI System Honk Powers Development

The transformation stems from Spotify's internal AI system Honk, a generative AI system that leverages Anthropic's Claude Code to handle coding tasks and accelerate product deployment.3

3

Honk enables remote, real-time code deployment using generative AI, fundamentally changing how engineers work. Söderström provided a concrete example: "An engineer at Spotify on their morning commute from Slack on their cell phone can tell Claude to fix a bug or add a new feature to the iOS app. And once Claude finishes that work, the engineer then gets a new version of the app, pushed to them on Slack on their phone, so that he can then merge it to production, all before they even arrive at the office."1

1

AI Does the Heavy Lifting for Product Velocity

Spotify credited the system with speeding up coding and deployment "tremendously," marking a significant leap in product velocity.1

1

The company isn't alone in this shift—Anthropic itself used Claude to build its recently launched Claude Cowork tool, while leaders at Meta and Microsoft have indicated AI is taking on increasing coding responsibilities.2

2

However, Spotify appears ahead of direct competitors like Apple Music and Amazon Music, which are still exploring AI integration in engineering. By building an AI layer fine-tuned to its own codebase, standards, and architecture, Spotify has created a competitive advantage that accelerates feature launches while maintaining quality.

Building a Unique Large-Scale Dataset for Personalization and Recommendations

Beyond coding efficiency, Spotify is developing a unique large-scale dataset that other LLMs cannot easily replicate. Gustav Söderström explained that music-related questions often lack factual answers—what constitutes workout music varies by geography and personal taste. Americans tend to prefer hip-hop overall, though millions prefer death metal, while many Europeans favor EDM and Scandinavians lean toward heavy metal.1

1

"This is a dataset that we are building right now that no one else is really building. It does not exist at this scale. And we see it improving every time we retrain our models," Söderström noted.3

3

This proprietary dataset offers Spotify an edge in personalization and recommendations that competitors cannot easily commoditize.

AI-Generated Music and Platform Monitoring

When analysts asked about AI-generated music, Spotify clarified its approach. The company allows artists and labels to indicate in track metadata whether a song was AI-assisted, but actively monitors the platform for spam content.3

3

This balanced approach enables innovation while maintaining platform quality. Söderström emphasized that the current state of AI development represents "not the end of the line in terms of AI development, just the beginning."1

1

The company's success with AI coding and accelerating product deployment suggests we're witnessing a fundamental shift in software development practices, with implications extending far beyond the music streaming industry.