From music app to everything app: Spotify’s relentless expansion
Spotify started as a music streaming app. Then it acquired podcast networks, signed exclusive deals with Joe Rogan and other marquee names, and repositioned itself as an audio platform. Then it added audiobooks, bundling them into premium subscriptions without asking users whether they wanted them. Now AI-generated content is the next layer in a pattern that has been building for over a decade.
Each addition has followed the same logic: expand the addressable market, increase time-in-app, justify the subscription price. The problem is that every expansion pulls the product further from the reason most people opened it in the first place — to listen to music. Users who signed up to stream albums now navigate an app cluttered with podcast recommendations, audiobook upsells, and an AI DJ that speaks between tracks.
The investor day announcement makes clear this is not a pilot program. Spotify’s leadership presented the AI push as a core strategic direction, the kind of commitment that gets built into product roadmaps and hiring plans for years. When a company announces AI strategy at the board level, it does not quietly walk it back when the feature gets mixed reviews.
What Spotify is building toward looks less like a music app and more like an AI-powered content engine — one that generates podcasts, music, and audio experiences rather than simply hosting them. That ambition puts it in competition with content studios, audiobook publishers, and AI media companies simultaneously. Whether Spotify can execute across all of those fronts is an open question. What is not open is the direction. The company has decided what it wants to be, and music streaming is now just the foundation, not the product.
The two kinds of AI Spotify could build — and which one it chose
Spotify faced a fork in the road, and the direction it chose matters more than most coverage admits.
On one side sits discovery AI — systems designed to analyze listening behavior, surface overlooked artists, and connect the right human-made music to the right ears at the right moment. Spotify already built a version of this with Discover Weekly and its DJ feature. That kind of AI makes human creators more visible. It treats the platform’s 100 million-plus track catalog as an asset worth excavating, and it puts listeners in front of work they would genuinely love but never find on their own.
On the other side sits generative AI — systems that produce the content itself. Music, podcasts, audiobooks, all manufactured on demand. Spotify announced at its investor day that it is pushing hard into this territory, piling AI-generation tools into the app across every format it supports.
These are not the same bet. Discovery AI and generative AI have opposite relationships to human creators. One amplifies them. The other competes with them.
Spotify chose the latter, and the consequences are already showing up. The platform faced public criticism for failing to properly label AI-generated music, eventually responding by adopting the DDEX industry standard for AI content tagging. That backlash did not slow the generative push — it just added a disclosure layer on top of it.
Most reporting frames this as a product story: Spotify is adding features. The more accurate frame is that this is a governance decision about whose interests a platform with half a billion users is designed to serve. A company that built its identity as the place where human creativity lives and earns a living is now building the infrastructure to replace that creativity with cheaper, faster, algorithmically produced substitutes.
That is a philosophical choice, not a feature rollout. And Spotify made it deliberately, at an investor day, in front of the audience whose approval it was seeking.
What AI-generated content actually means for the creators on the platform
Spotify built its entire catalogue on the backs of human creators. Every song, podcast episode, and audiobook on the platform was made by a person — often an independent artist or creator who relied on Spotify’s reach to build an audience and generate income. That foundational relationship is now under direct pressure.
By rolling out AI tools capable of generating music, podcast-style audio, and spoken-word content, Spotify is no longer just a distributor for human work. It is becoming a producer of content that competes with the very creators who filled its library in the first place. An independent musician trying to land playlist placements now exists on the same platform as AI-generated tracks that can be produced at a fraction of the cost and in a fraction of the time. The same dynamic applies to podcasters and audiobook narrators.
The royalty question sits at the center of this shift, and Spotify has not answered it. The company’s public communications around its AI content push contain no clear framework for how — or whether — AI-generated content generates royalties, who receives them, and how those payments compare to what human creators earn per stream. That silence matters. Spotify has faced years of criticism from artists over low per-stream payouts, and introducing a new category of content with no stated compensation structure deepens that tension.
There is also a labeling problem with a recent history. Spotify faced public backlash after failing to properly identify AI-generated music in its catalogue, and only updated its policy afterward, adopting the DDEX industry standard for disclosure. That the policy change came after criticism — not before — signals that creator protections are reactive rather than built into the platform’s AI strategy from the start. For the musicians, podcasters, and narrators who treated Spotify as a primary revenue channel, that sequencing is not reassuring.
The listener experience problem: more content, harder choices
Spotify already hosts over 100 million tracks, tens of millions of podcast episodes, and a growing audiobooks catalogue. Finding something genuinely worth your time inside that catalogue is already a friction-filled experience for most users. Flooding that same library with AI-generated music, narration, and podcast content does not solve the discovery problem — it compounds it.
The labelling issue sharpens that risk. Spotify faced direct criticism for failing to properly identify AI-generated music on its platform and subsequently adopted the DDEX industry standard for flagging such content. But a policy change and actual enforcement are two different things. AI can now produce tracks faster than Spotify can process and categorise them. If listeners routinely encounter AI-generated content without clear, prominent disclosure, the implicit contract between the platform and its audience — that you are listening to a human being who made something — quietly breaks down. Trust, once eroded, does not recover through a policy update.
The deeper irony is what Spotify is choosing not to prioritise. The feature most associated with the platform’s identity — personalised recommendation, the engine behind Discover Weekly and Daily Mixes — is precisely the kind of AI investment being pushed to the back of the queue. These tools use listening behaviour to surface music users would not have found on their own. They represent AI working in service of the listener. Spotify’s current strategic direction points the other way: AI that generates content for the platform’s benefit, not tools that filter and curate on behalf of the person with the headphones in.
More content without better discovery means listeners spend more time scrolling and less time actually listening to things they love. That is a product failure regardless of whether the tracks are made by humans or machines.
The business logic: why AI-generated content is attractive to Spotify’s bottom line
Spotify’s relationship with human-created content has always been expensive. The company pays licensing fees to major labels, negotiates royalty structures with publishers, and splits revenue with rights holders across music, podcasts, and audiobooks. Those obligations eat directly into margins. AI-generated content changes that equation entirely — Spotify can own it outright, pay no royalties, and avoid the complex multi-party negotiations that have defined its business since launch.
This is the margin story that investors actually care about. Spotify has spent years operating as a distributor of other people’s work, which caps how much of each dollar it can keep. A platform that generates its own content — and owns the intellectual property — operates under a fundamentally different financial model. The shift from distributor to owner is not incremental; it restructures the entire cost base.
Read through that lens, Spotify’s investor day announcement looks less like a product vision and more like a balance sheet strategy. The company framed its AI push around improving the listening experience and giving creators new tools. The underlying logic points elsewhere. When Spotify generates an AI ambient music track or an AI-voiced podcast, it keeps revenue that would otherwise flow to a label, a licensor, or an independent artist. Multiply that across millions of streams and the financial impact becomes significant.
Spotify has already demonstrated willingness to use its platform to favor cheaper content. The company faced sharp criticism for flooding its catalog with AI-generated tracks, several of which surfaced on curated playlists without clear labeling. That controversy forced a policy change, but it also confirmed that AI content was already moving through Spotify’s systems at scale — and that Spotify was not rushing to slow it down. The business incentive to keep that content flowing is direct and measurable. The incentive to prioritize human creators, whose work comes with licensing costs attached, runs in the opposite direction.
What to watch: the signals that will tell us how this really plays out
Three indicators will determine whether Spotify’s AI push reshapes audio culture or simply inflates the platform’s catalogue numbers.
The first is labelling. Spotify faced direct criticism last year for failing to properly identify AI-generated music, and only adopted the DDEX industry standard after that backlash forced its hand. Whether the company applies that standard consistently across music, podcasts, and AI-generated audiobooks — or quietly lets it slide as volume scales — signals how much Spotify actually respects listener choice. A platform comfortable with users not knowing what they are hearing is a platform that has already made its decision.
The second is creator response. Musicians, podcasters, and voice artists are not a passive constituency. Visual artists and writers built coordinated resistance movements when AI-generated work began flooding their professional spaces, and audio creators are watching the same pattern unfold. If AI-generated tracks start appearing in curated editorial playlists alongside human work — without clear separation — that friction escalates fast. Industry bodies, union agreements, and public pressure campaigns follow. Spotify has the catalogue leverage to absorb short-term noise, but sustained creator withdrawal or legal challenges would stress-test that position.
The third, and most revealing, indicator is recommendation quality. Spotify’s core product promise has always been helping listeners find audio they actually want. Its investor day announcements leaned heavily toward using AI to generate content rather than to improve discovery. As AI-produced tracks multiply across the catalogue, the recommendation engine faces a harder signal-to-noise problem. If Spotify’s algorithms start surfacing AI filler because it is cheaper to license and easier to produce at volume, engagement metrics will eventually reflect that — and subscriber retention will follow. That outcome benefits no one except the cost side of a balance sheet.
Watch those three data points. They will show, faster than any press release, who this strategy was actually designed to serve.
