Spotify says its future depends on being the music service AI agents choose

Spotify closed its third quarter with 788 million monthly active users and nearly €5 billion (around $5.6 billion) in revenue, according to TechCrunch — but the headline numbers were not the most interesting part of the earnings call. The notable signal was strategic: Spotify is positioning itself as the music service that AI agents reach for when a user asks for something to listen to.

Entertainment as an agent-proof category

Co-CEO Gustav Söderström told investors that Spotify is one of the few businesses likely to benefit as AI agents spread, because agents are best at taking over chores people would rather avoid. Entertainment works differently, he argued: people want the listening itself, not a condensed version of it. As TechCrunch quotes him, "Nobody asks an agent to 'summarize a song' for them."

The theory is that listeners will delegate the logistics — finding music for a run, tracking down a podcast about a topic they were just discussing with an assistant, or resuming an audiobook exactly where they left off — while the actual listening still lands on Spotify.

Integrations with ChatGPT, Claude and Meta's Muse

Spotify is already acting on that bet. According to TechCrunch, it was the first music service to integrate with Meta's Muse assistant, and it also works with ChatGPT and Claude. Söderström framed the company's goal as becoming the media service people use from any agent, turning the growth of agents into a tailwind rather than a headwind, and said Spotify is seeing strong adoption inside these assistants.

That framing is a deliberate contrast with much of the rest of the industry.

A market split over agent access

Friction is already emerging over how agents reach third-party services. TechCrunch points to Amazon, which has blocked Meta's Muse from shopping on its platform — an early example of a major player walling itself off over questions of access, security, privacy or commercial control. Söderström contrasted that posture with Spotify's, describing the company as "leaning in" to agent growth while others try to figure out their response.

Personalization as the moat

Spotify's second argument is that an agent-mediated world rewards services that already know the user. Söderström described Spotify as an intelligent media service wrapped in an agent — something you can talk to, question and instruct, whether inside Spotify's own app or through another assistant. One example he gave: asking Spotify to comb through a listening history stretching back to 2008, pull out a listener's genuine favorites, speed them up or slow them down, sort them by key and BPM, and blend them into a running playlist matched to the runner's cadence.

A generic assistant without that history would produce a far blander playlist. Spotify has years of data on which artists people return to, which songs they skip and what they discover.

Notably, that data stays home. Söderström said Spotify does not send listener data to its partners; instead, an agent passes along a request, and Spotify's own systems reason about it and decide what to play.

Why it matters

Spotify's pitch is one of the clearest statements yet of how consumer platforms might adapt to agent interfaces. If asking an assistant becomes a default way people find media, distribution shifts from app icons and home screens to whichever services agents are wired into — and being early and welcoming becomes a competitive advantage, much as app-store placement once was. It also sketches the emerging divide in the market: some platforms, like Amazon in the Muse case, are restricting agents, while others are inviting them in. And Spotify's privacy architecture — keeping behavioral data in-house while exposing an interface that agents can call — points at a likely template for how incumbents with rich user data can join the agent era without giving away the asset that makes them valuable. The open question is how much of the listening experience agents will actually mediate, and whether being the default music backend for assistants proves to be a durable position or a commoditized one.