Spotify said its top developers haven’t written a single line of code since December, thanks to AI.

Has AI coding reached a tipping point? At least that seems to be the case with Spotify. On its fourth-quarter earnings call this week, the company’s top developers said they “haven’t written a single line of code since December.” This statement from Spotify co-CEO Gustav Söderström came alongside other comments about how the company is using AI to accelerate development.

For reference, Spotify says it’s rolling out more than 50 new features and changes to its streaming app throughout 2025. And most recently, we’ve launched more features like AI-powered prompted playlists, page matching for audiobooks, and About This Song, all released within the past few weeks.

Spotify’s engineers told analysts they are using an internal system called “Honk” to speed up coding and product. This system allows for things like generative AI, especially remote real-time code deployment using Claude Code.

“As a concrete example, on the morning commute, an engineer at Spotify might use his phone in Slack to tell Claude to fix a bug or add a new feature to the iOS app,” Söderström said. “And when Claude is done, engineers can get the new version of the app and push it to Slack on their phones to merge it into production before it even gets to the office.”

Spotify said the system helped speed up coding and distribution “tremendously.”

“We anticipate that this is only the beginning, not the end, in terms of AI development,” Söderström said.

He also highlighted Spotify’s ability to build unique data sets that other LLMs cannot commercialize, like other online resources such as Wikipedia. That’s because questions about music don’t always have factual answers, he said.

For example, if you ask people what exercise music is, you’ll sometimes get different answers depending on the region. Americans tend to prefer hip hop overall, but millions prefer death metal. And while many Europeans like EDM, many Scandinavians like heavy metal.

“This is actually a data set that no one else has built – it doesn’t exist at this scale – and we see improvements every time we retrain the model,” Söderström said.

Analysts on the call also asked about Spotify’s approach to AI-generated music. The company explained that while it allows artists and labels to indicate in a track’s metadata how a song was created, it is still cracking down on the platform against spam.