Spotify has revealed that its top developers have not written a single line of code since December, a striking indication of how deeply artificial intelligence is transforming its engineering workflow. The revelation came during the company’s fourth-quarter earnings call, where co-CEO Gustav Söderström highlighted the accelerating role of AI in software development. Spotify shipped more than 50 new features and updates to its music streaming platform throughout 2025, including recent launches like AI-powered Prompted Playlists, Page Match for audiobooks, and About This Song—each introduced in the past few weeks. These rapid advancements are being driven by an internal AI system called “Honk,” which enables engineers to deploy code remotely and in real time using generative AI, particularly through integration with Claude Code. Söderström described a scenario where an engineer on their morning commute could use Slack on their phone to ask Claude to fix a bug or add a new feature to the iOS app. Once the AI completes the task, the updated version of the app is pushed directly to the engineer’s phone via Slack, allowing them to review, test, and merge the changes into production before even arriving at the office. “We’ve seen this dramatically increase our coding and deployment velocity,” Söderström said, adding that this is not the end of the AI transformation, but rather the beginning. Beyond speed, Spotify emphasized the uniqueness of its data. Unlike other online resources such as Wikipedia, which can be easily replicated or commoditized, Spotify is building a vast, nuanced dataset around music preferences and behaviors—information that lacks universal factual answers. For example, responses to “What is workout music?” vary widely based on geography and personal taste: Americans often lean toward hip-hop, while others prefer death metal; many Europeans enjoy EDM, but Scandinavian listeners frequently favor heavy metal. This rich, context-sensitive data, gathered at scale, gives Spotify a competitive edge in training models that understand music in human terms. Söderström noted that this dataset is continuously improving with each retraining of Spotify’s AI models. The company also addressed concerns around AI-generated music, explaining that it allows artists and labels to specify in a track’s metadata how a song was created. Spotify remains vigilant in monitoring its platform for spam and ensuring authenticity, balancing innovation with quality control.