B Hari

October 22, 2026

A Human Musician Still Needs to Be Heard


Protecting music from AI spam is necessary. The harder test is whether unfamiliar human artists can still find an audience.


On October 20, Spotify said it had excluded more than 50 million tracks classified as slop or spam from recommendations this year. That is a company report about filtering, not independent evidence that musicians are earning more. The distinction matters: a cleaner music service can still leave most musicians unheard. Protecting human artistry should include a credible chance to find listeners, alongside protection from synthetic clutter.


The immediate benefits are real possibilities. Listeners should encounter fewer deceptive uploads; musicians should face less competition from industrial quantities of unwanted material. Spotify also describes credits disclosing creative uses of AI and tools that let artists review releases attributed to them. These measures recognise that assistance, impersonation and spam are different things. The company acknowledges that AI detection has limitations.


But authenticity answers only one question: who is behind the recording? It does not answer who gets heard.


Imagine a singer releasing her first songs in a language with a small commercial market. This is a hypothetical artist, not a documented victim of Spotify's changes. She may welcome help with production and protection against imitation. Yet a badge confirming her humanity will accomplish little if the service rarely introduces her to anyone. Her scarce resource is a first serious hearing.


This problem predates generative AI. The UK's competition authority reported in 2022 that the top 0.4% of artists accounted for more than 60% of streams. It also found that digitisation benefited consumers and opened routes for artists, and did not identify ineffective competition as the explanation for creators' income concerns. Those historical UK findings are not an estimate of today's Indian market. They show why wider access and unequal rewards can coexist.


Cheaper production intensifies that tension. If making another recording costs less, supply can grow faster than anyone's listening day. The economic prize moves towards finding, recommending and sustaining attention. Platforms can gain loyalty from reliable discovery; established catalogues may capture attention freed by spam removal. Neither mechanism guarantees that a newcomer receives a larger income. Generating more songs is a productivity measure with little meaning unless someone values hearing them.


Musicians also care about how they appear. A 2021 interview study by Andres Ferraro and colleagues explored concerns about reaching audiences, transparency and the loss of context around music. Its nine interviewees cannot represent all artists. Their accounts nevertheless illuminate something listening totals miss: a musician is building a body of work and a relationship, not merely supplying interchangeable audio.


India adds another constraint. An EY–Indian Music Industry study published in July, surveying 15,373 smartphone users, found that only 38% said they had ever paid for music streaming, including bundles. This is industry-linked survey evidence, not a census or a count of current subscribers. It cautions against assuming that more listening automatically produces much more money to share. Expensive verification or promotion could therefore burden precisely the artists with the least financial room.


UNESCO's 2022 analysis
identified both digital-access barriers and business models that hinder discovery of diverse cultural expression. For a platform serving many languages, success should consequently be tested across those languages. A smooth experience for dominant audiences cannot establish that smaller musical communities have gained meaningful access.


There is a serious objection: listeners want enjoyable music, not a cultural policy imposed through their headphones. Stronger demands on platforms could also delay useful protection. And affordable AI tools may let people create who could never hire a studio. These arguments favour clear disclosure, effective anti-spam enforcement and listener choice. They do not require treating existing popularity as the limit of what people might enjoy.


A better future would combine cheaper creation with trusted credits and deliberate opportunities to discover unfamiliar artists. An uneven future would give established musicians cleaner profiles while leaving newcomers invisible. A harmful one could reward mass production, or make proving legitimacy so burdensome that small human acts withdraw. These are possibilities shaped by design and funding, not predictions about Spotify.


Questions for founders and investors


Does the product help unfamiliar artists build returning audiences, or simply keep familiar songs playing? Who pays for documentation and mistakes? What happens to discovery and net creator income across languages and audience sizes?


Platforms should publish those outcomes alongside enforcement totals, fund accessible artist support, and offer listeners meaningful discovery controls. Cultural institutions can commission independent evaluations and support local curators. Protecting human music should leave room for the human musician nobody knows yet.


Sources