Spotify Goes All In On AI Coding, Says Co-CEO

Photo Credit: Emilipothèse

Spotify shared during its Q4 earnings call that the company’s developers have relied on AI to write the lion’s share of its code since December.

Spotify held its fourth-quarter earnings call this week, revealing the many ways the company has been relying on artificial intelligence to accelerate the platform’s development. According to Spotify co-CEO Gustav Söderström, the company’s top developers “have not written a single line of code since December.”

The Stockholm-based streaming giant has rolled out over 50 new features and changes to its mobile app over the past year, including its recently released AI-powered Prompted Playlists, audiobook Page Match, and About This Song.

Söderström noted this rapid rollout has been largely thanks to an internal system used by the company’s engineers called “Honk” in order to speed up coding and product velocity. The system enables engineers to deploy code remotely and in real-time using generative AI provided by Claude Code.

“As 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,” said Söderström. “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.”

He also noted that Spotify has the ability to build a unique dataset that other large language models (LLMs) could not capitalize on in the way they could other online resources, such as Wikipedia. This, he said, is because music-related questions are often fluid and heavily based on region.

For example, people from different parts of the world may have varying concepts of what constitutes “workout music.” Americans might prefer hip-hop, and many Europeans might prefer EDM, while Scandinavians might prefer to work out to heavy metal. Söderström said this is the type of dataset that Spotify is building, which sets it apart—because the company has the largest swath of users from which to pull data.

“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,” he said.

The company also clarified its approach to AI-generated music, explaining that artists and labels can indicate in a track’s metadata how the song was made and the extent to which AI was used. However, Spotify noted that the platform is still “policing” for AI-generated slop tracks.