Spotify Bets on AI Agents to Become Music's Default Choice

At a time when tech companies are racing to understand where they stand in the wave of AI agents, Spotify has taken a clear position: rather than resisting this wave or waiting on the sidelines, it has decided to become part of it. During its third-quarter earnings call, the most notable figure wasn't the 788 million monthly active users, nor the revenues approaching €5 billion (around $5.6 billion), but rather the vision the company laid out for how people will listen to music in the years ahead.

Why the Co-CEO Believes Entertainment Is Immune

Gustav Söderström, the company's co-CEO, presented a striking argument to investors: AI agents excel at handling the tasks humans dislike and would rather offload — summarizing documents, scheduling appointments, or completing routine work. But entertainment operates on entirely the opposite logic; no one asks a smart assistant to "summarize a song for them."

Söderström's conclusion is that music, podcasts, and audiobooks belong to one of the few categories that could genuinely thrive in the age of agents rather than suffer from it. An agent doesn't consume entertainment on the user's behalf; it helps them reach it faster and more precisely.

From an App to a Service Summoned from Anywhere

The shift Spotify is describing changes the very point of entry. Instead of the user opening the app, searching for an artist, and building a playlist themselves, they may simply issue a conversational request to a smart assistant, such as:

• Music suited to a run at a specific tempo.
• A podcast discussing a topic the user was talking about with the assistant moments earlier.
• A summary of an audiobook picking up exactly where they left off.

And here the company's goal comes into focus: to be the service these agents turn to when executing such requests. Spotify was the first music service to integrate with Meta's "Muse" assistant, alongside its integrations with ChatGPT and Claude.

Söderström says the goal is to "turn the growth of agents into a tailwind rather than a headwind," adding that Spotify usage rates from within these agents are encouraging, and that the company is choosing to participate in the growth rather than stand on the sidelines watching it — at a time when other companies remain hesitant or are pulling back.

The Access Battle: Friction Has Already Begun

This openness is not the market norm. The outlines of a conflict over agents' right to access third-party services have already begun to emerge, with one of the clearest examples being Amazon's blocking of the "Muse" assistant from making purchases on its platform. This example reveals that major players may close their doors to external agents when questions arise around commercial control, privacy, security, or ownership of the customer relationship. Spotify's choice of early openness gives it a different negotiating position: to be the default audio destination within the agent ecosystem.

The Real Advantage: Years of Listening History

The company today describes itself as "an intelligent media service wrapped in the form of an agent," whether accessed directly through its app or via an external assistant. According to Söderström, Spotify is no longer merely an interface with buttons to press; it has become a party you can converse with, question, and instruct.

The example he offered illustrates the idea: a user could ask Spotify to review their listening history song by song going back to 2008, extract their genuine favorites, adjust their tempo by speeding them up or slowing them down, then order them by key and beats per minute, and combine them into a running playlist that precisely matches their stride.

This capability rests on an asset no general-purpose assistant possesses: a long behavioral record that knows which artists the user returns to, which songs they skip, and which patterns they discover. That's what makes a recommendation far more personal than a generic list generated by an assistant with no knowledge of its owner's taste.

Privacy: The Request Is Sent, Not the Data

Söderström emphasized a fundamental point in this equation: Spotify does not share its listeners' data with partner companies. Under the adopted model, the external agent sends only a request, after which Spotify's internal systems analyze that request and decide what to play. In this way, the company keeps its most valuable asset within its own boundaries, selling the outcome rather than the raw material.

The Practical Lesson for Businesses

What Spotify is doing offers a useful framework for any digital organization: the question is no longer "Will AI replace my service?" but rather "Will agents find my service callable and usable?" Those who own deep behavioral data and open, integration-ready interfaces are more likely to gain from this shift than lose to it.

✦ بقلم فريق دروب أيديا