Blog

Athens Music Week runs 6-9 May 2026. Goethe-Institut Athens for the conference, PLEX, ΚΛΑΚΑΖ and the French Institute for showcases. The 8th edition is held under the auspices of the Hellenic Copyright Organization (HCO).

Over 100 activities on the programme. Most aren't for you. The 14 below are. If you run an indie label or work on the team, this is your AMW. Sessions marked Closed require an invitation or are members-only. Sessions marked Registration required are open but capped, so book now.

Wednesday 6 May

15:45 - Bridging Balkan Borders listening session: New music from Southeast Europe | Goethe Auditorium

A first listen to material from a 4-day SE European songwriting residency. In the room: Konstantinos Adam (A&R Indie Pop & Alternative, Stay Independent), Vanias Apergis (Founder & Director, Yellowgroove), Marko Anastasov (CEO, MUZE Corp / Bulgaria), Dimitris Beltsos (Songwriter/CEO, Arcade Music) and Jay Stolar (Executive Producer, FATE RECORDS & HUME / US).

If your A&R radar covers the SE European corridor, the first concrete read of the week.

18:00 - AMW Reception by EKKOMEΔ | Goethe Terrace

The Hellenic Film and Audiovisual Center (EKKOMEΔ) hosts the opening reception. Music and audiovisual professionals in one room with the broader creative industries. Registration required.

If you've got sync ambitions or any catalogue exposure to film and audiovisual, the most efficient first-night networking moment of the week.

Thursday 7 May

11:30 - Beyond the hubs: The rise of Europe's next touring markets | Goethe Seminar Room

Christian Cléret (CEO, Dionysiac Tour) and Leonardo Lupelli (Federation of Music Conferences) on Southeast Europe, the Balkans and the Mediterranean as live markets. Niko Michault (P.U.S.H Music Management), Mark Dieler (Sicily Music Conference) and artist Melina Vlachos round out the room. The corridor AMW is built around. Question after the panel: who in your team owns the territory plan?

12:00 - AI & music: Moving past the fear, towards real practice | Goethe 2004

Matt Cartmell (Chief Executive, Music Technology UK) moderates, with Pantelis Vikatos (Sr Director Research & Innovation, Orfium), Thais Ruiz de Alda (Digitalfems) and Stelios Karozis (Pharos AI Factory). What label operators are doing now with AI tools, what's integrated, what isn't. If you haven't formed a position on AI tooling yet, treat this as a briefing.

12:00 - Ο ρόλος του Εκδότη / The Role of the Music Publisher | Goethe Auditorium

Greek-language panel running parallel to the AI session. Vanias Apergis (Yellowgroove) moderates, with George Myzalis (COO, EDEM Rights), Ilianna Antoniou (AUTODIA), Stefanos Dasopoulos (Stay Independent), Andreas Papadopoulos (D-Version Music Publishing) and Irina Loukatou (Musou Music Group). Stay Independent + AUTODIA + EDEM Rights together is the most concentrated Greek publishing-and-CMO conversation of the week. You'll need to pick: AI room or this one.

13:45 - Built for music: Tools, ideas and real-world solutions | Goethe 2004

A live show-and-tell of music business tools. Hosted by Becky Brook and Sakis Triantafyllakis (Director of Operational Excellence, Orfium). Founders demoing include Pierre-Alban Mulliez (CEO, Claimy), Frédéric Herout (CEO, Music Magnet), Benoit Menet (CEO, Music Tomorrow), Gérôme Vanherf (CEO, ikii), Dimitris Komninos (CEO, Mood) and Lazaros Penteridis (CEO, ComeTogether).

For label ops leads still on disconnected spreadsheets and platform dashboards, the most efficient hour of the week to scan what's available.

15:00 - From local traction to international markets: How Greek artists grow | Goethe Seminar Room

The flagship Thursday panel for indies. Powered by The Orchard. Mandy Salem-Aubry (International Conference Director, AMW) moderates, with Mihalis Kaloudis (Director, The Orchard Greece), Andreas Periklis Metaxas (Stay Independent), Andreas Kefalas (CDO / Head of Copyright, Broma 16) and Panagiotis Loulourgas (Head of International / A&R Manager, Panik Records). The senior people running the strongest international arms in the Greek market on one stage.

Friday 8 May

11:30 - Spotify Masterclass: Growing your audience and reaching beyond borders | Goethe 2004

Spotify's Southern & Eastern Europe team (Emilia Mihailescu, Katarzyna Czechowska, Ekaterina Pirogova) on Spotify for Artists tools, pitch flow and audience growth. Regional team, regional context. Registration required.

Come with one specific catalogue question. The Q&A is worth more than the slides.

13:15 - Greece & Bulgaria in the new Southeast European music landscape | Goethe 2004

Powered by BAMP. Our CEO Luciano Winter is a speaker on this panel.

Moderated by Boral Shen (Executive Director, Bulgarian Association of Music Producers), with Manos Dedevesis (COO & Head of Publishing, Stay Independent), Stanislava Armoutlieva (CEO/Owner, Virginia Records), Magdalena Sotirova-Ivanova (Managing Partner, Monte Productions) and Oleg Rozov (Director Eastern Europe & Central Asia, The Orchard).

The two gateway markets into a region of over 70 million people (Serbia, North Macedonia, Turkey). Schengen developments matter operationally for live touring and crew mobility. If you're active or planning activity in SE Europe, this is where the operating reality gets discussed by the people running the labels.

14:30 - Innovation and the law: Where to move fast and when to stop | Goethe 2004

Mandy Salem-Aubry moderates, with Eric Reithler-Barros (CEO, Fold Artists), Becky Brook, Chris Hocking (Founder, The Factory Music Group), Eva Moustaka (HCO) and Alexandros Nousias (Pharos AI Factory). Suno, Udio and training-data attribution under existing copyright frameworks. Are independent rights holders in the rooms where licensing terms get set?

16:15 - How to build a direct fanbase that lasts | Goethe 2005

Marion Sosnowicz (Head of Business Development, Music Glue) on direct-to-fan operations at scale. Iron Maiden, James Blunt, Enter Shikari in the back catalogue of campaigns. 60-minute interactive workshop. Registration required.

If you're reducing platform dependency on streaming, where the operating model gets concrete.

Saturday 9 May

11:00 - HERMES Association inaugural event | Conference Goethe Auditorium [Invite only]

The official launch of HERMES, the new Association of Greek Independent Music Companies. Hosted by Manos Dedevesis (COO & Head of Publishing, Stay Independent).

The single most significant moment of AMW 2026 for European indies. A new national indie association in a market where Stay Independent currently runs up to 58% of the local digital chart share has direct implications for how Greek catalogue value gets collectively represented at European policy level. If you do business in or with Greece, be in the room.

13:45 - AI, copyright and Greece: The rights question no one can avoid | Goethe 4004

Yannis Iliopoulos (Senior Programmes Manager, ICMP London / Founder, The Sound of Everything) moderates, with Maria-Daphne Papadopoulou (Acting Director, Hellenic Copyright Organization) and Maria G. Sinanidou (Counselor at Law / Certified Mediator, HCO).

The Acting Director of the HCO sitting on a public AI panel is the signal. EU AI Act, training data, transparency requirements as they're being interpreted in Greece right now. For label heads with active songwriter relationships and catalogue exposure, this connects the morning's HERMES launch and Friday's Innovation and the Law panel into one operational picture.

15:00 - Beyond the Algorithm: Curators, Collectors & Community | Goethe 3002

Powered by Crates. Stathis Kalatzis moderates, with Matthias Strobel Hohmann (President, Music Tech Germany), George Ilias (Founder, Crates), Katerina Mortzou (DJ, K.atou) and Vassilis Sevdalis.

Where the value of taste gets captured commercially when the algorithms keep flattening it. If you invest in catalogue depth and curation, this one.

Saturday closes with the strongest showcase curation of the week: Lefty (GR), Foehn (FR), Rhumba Club (UK), 505 (IT/DE), Hana Fatur (SI), Obadu (HU), Iota (GR).

Closed-door sessions worth knowing about

Two sessions on the programme are invitation-only and won't open to general delegates. Worth tracking because the conversation shapes what surfaces in the public panels:

• Thu 16:00 - Music Tech Europe General Assembly | Goethe 3002. Members and partners only, classified as a Private Meeting on the AMW programme. Hosted by Jose Zagazeta (Barcelona Music Tech Hub) and Turo Pekari (Copyright Delta), with Anna Zò (President, Music Tech Europe), Georges Perot (Artistic Director, AMW), Matthias Strobel Hohmann (Music Tech Germany) and Matt Cartmell (Music Technology UK).
  
  

• Fri 10:30 - Music & AI Forum "Knowledge as Infrastructure" | Goethe-Institut Athens. Closed expert roundtable hosted by Music Tech Europe. Moderated by Yvan Boudillet (Secretary General, Music Tech Europe) and Turo Pekari, with Konstantinos Karavezis (CMO & CFO, Stay Independent), Eric Reithler-Barros (Fold Artists), Henry Umunnakwe (Metavallon VC) and Michael Petychakis (CTO, Dialectica). Output from this room will surface in Friday afternoon's Innovation and the Law panel.

The pattern across four days

HERMES launches Greece's first dedicated indie association. The southeast European corridor gets mapped as a live and catalogue-value region of Europe. AI licensing frameworks are being written, mostly without independent rights holders in the room.

Every session above maps back to one question: who's building the infrastructure that the next decade of European independent music will run on, and who's paying rent to use it.

What happens after the summit is decided by the operators paying attention.

What the algorithm looks for is more material to serve. An artist with only a handful of releases gives the algorithm very little to work with. An artist with a substantial back catalogue, built consistently over time with strong engagement patterns, gives the algorithm a rich picture of who this artist is and who their audience is. That picture is what enables the recommendation engine to surface your music confidently to new listeners who share the taste profile of your existing fans.

This is the compounding dynamic that matters most for independent labels. Spotify's recommendation engine pulls from three signal sources: collaborative filtering (who listens to what alongside what), natural language processing (what people write about your tracks across the web), and raw audio analysis (what your music sounds like to a model trained on millions of tracks). Every release contributes to all three. When someone who likes Artist A also saves or replays a song by Artist B, the model strengthens that connection. Every new track adds another audio fingerprint and another set of contextual references. The more material an artist has, the more of those signals accumulate, and the more confidently the algorithm can recommend them to the right listeners. A deep catalogue is the dataset the algorithm uses to understand and distribute your music.

Spotify's own Loud and Clear data supports this. According to the 2025 report covering 2024, nearly a quarter of the 12,500 artists who generated over $100,000 in royalties that year had not been releasing music professionally five years earlier. The most recent report, published in March 2026 and covering 2025, puts that figure at more than 13,800 artists. Success in streaming is built on accumulation: more releases, more data, more algorithmic confidence, wider reach.

The perfection trap

The creative instinct to release only when something is ready is understandable. Releasing imperfect work feels like a professional risk. But the streaming economy does not penalise an average track the way a physical release cycle used to. A song that finds its audience is a success regardless of when it was created or how long it took. A song that never comes out contributes nothing to your catalogue, your algorithmic profile, or your audience's relationship with your music.

The more common cost of waiting is a data cost. Every month without a release is a month without new engagement signals entering the system, a month without Release Radar appearances reaching your existing followers, a month without Discover Weekly placements introducing your music to listeners with adjacent taste profiles, and a month without the algorithm refining its model of who your audience is. A six-month gap between releases weakens the data signal Spotify is working with. Rebuilding it from that weakened state takes time.

Quality still matters. There is a real difference between an average track and a track listeners actively reject. An average track sits in the catalogue, accumulates modest engagement, and feeds the model. A track that listeners abandon inside the first 30 seconds, the threshold at which a play registers as a stream rather than a skip, sends a negative signal that works against your algorithmic standing. Skip rate and save rate are core inputs to the recommendation engine. The case here is for releasing consistently rather than holding everything back for a moment of maximum readiness that may never arrive, and that the algorithm rewards far less than it rewards sustained presence.

What a consistent release strategy looks like in practice

The cadence most industry practitioners point to sits somewhere between four and eight weeks per release. Frequent enough to maintain a consistent presence in Release Radar, regular enough to keep accumulating algorithmic signals, and spaced enough that each release gets its own promotional window before the next one arrives.

The International Music Summit runs April 22 to 24 in Ibiza, hosted at Mondrian and Hyde. Ibiza in Cala Llonga. The theme this year is “Reclaim The Dancefloor.” The framing is not accidental. It signals that the people who built the independent electronic music sector are being asked, directly, whether they still control the conditions they built it in. For independent labels, distributors, and professional artists operating in Europe, IMS is three days of concentrated signal about where the business is heading. What follows is a session-by-session guide to the programme moments that matter most for your operation.

Wednesday 22 April

12:05 - IMS Business Report 2026 | Summit Stage Indoor

Every year, IMS opens with the Business Report presented by MIDiA Research founder Mark Mulligan. This is the annual industry valuation: a full analysis of cultural, financial, and economic trends shaping the global electronic music market. For distributors and label services companies, this document benchmarks what the market is actually worth and how revenue is flowing through the ecosystem. It is the clearest starting point available for any strategic conversation about the next twelve months.

12:35 - IMS Business Report Analysis | Summit Stage Indoor

The analysis that immediately follows brings Dan Chalmers (YouTube Music EMEA, Google), Finlay Johnson (CEO, AFEM), and Mark Mulligan back to the stage to translate the numbers into decisions. Streaming economics, market consolidation, and where independent labels sit within that picture are all on the table. This session answers the question the report raises: so what do we do with this?

16:15 - Get Played, Get Paid: Ibiza Club Focus Revisited | Summit Stage Outdoor

Hosted by AFEM, this session has been running for over a decade. The premise is straightforward: copyright societies should use modern tracking technology to ensureaccurate distribution of license fees from clubs and festivals to the music creators and rights holders entitled to them. The panel includes Yuri Dokter (CEO and founder of KUVO Powered by DJ Monitor, based in the Netherlands) and representatives from BMAT and SGAE, moderated by AFEM co-founder Kurosh Nasseri. For European independent labels and rights administrators, this is a direct progress check. The technology exists. The question this session addresses is whether the institutions responsible for collecting and distributing these fees have actually moved.

17:00 - Keynote Interview: Yann Pissenem | Summit Stage Indoor

Founder and CEO of The Night League and Ushuaïa Entertainment, Yann Pissenem built some of the most commercially successful nightlife venues in the world while maintaining a clear creative and cultural point of view. The session, moderated by Pete Tong, is less about the venues themselves and more about the architecture of long-term independent decision making at scale. For anyone building an independent business in electronic music, this conversation is worth the time.

16:55 - Eastern Europe on the Rise: New Voices, New Futures | Summit Stage Outdoor

Presented by Telekom Electronic Beats, this panel brings together festival directors, label executives, and label owners from Poland, Hungary, Germany, and the wider CEE region to map how Eastern Europe has developed its own ecosystem over the last decade. The conversation includes Basia Klaczak (Unsound, Poland), Laszlo Papp (Inota Festival and Electronic Beats, Hungary), and Marie Montexier (label owner, Germany). For European independent labels and distributors thinking about territory expansion and A&R strategy, this session names what is already visible in the streaming data: the next wave of catalogue value in European electronic music is being built outside the traditional market centres.

Thursday 23 April

12:00 - The AI Divide: What Authenticity Means to Fans Now | Summit Stage Indoor

Deezer’s CMO Maria Garrido presents findings from a study conducted by Deezer and Ipsos across 9,000 people in eight countries on how listeners perceive fully AI-generated music. The implications are direct for labels and distributors managing catalogue decisions and release strategy. Fan trust is not a soft metric. It determines whether a release earns algorithmic momentum or disappears quietly. This session gives that question a data-backed answer.

12:15 - Mathame Presents NEO: A World Where Humans and AI Coexist | Summit

Stage IndoorImmediately following, Italian artist Mathame and his brother Riccardo Giovanelli present NEO, the first large-scale AI awareness project operating at the centre of entertainment. It is notable not just for what it is, but for what it signals about how forward-thinking independent artists are choosing to position themselves in relation to AI rather than waiting for the industry to resolve the question for them.

12:25 - AFEM x AIxchange: Shaping New Revenue Streams in the Age of Generative AI | Summit Stage Indoor

Ralph Boege (CEO, Paradise Worldwide) presents AIxchange, a consent-based licensing framework for recordings, compositions, and metadata used in AI training. The framework is grounded in transparency, attribution, and fair compensation, building directly on the AFEM AI Principles. For independent labels with deep catalogues, this is not an abstract policy debate. AI companies are training on music right now. The question is whether independent rights holders are in the room where those licensing terms are being set.

12:35 - AI: Inside the Moral Revolution | Summit Stage Indoor

The AFEM-hosted panel that ties all three AI sessions together. Participants include Aly Gillani (Artist and Label Relations Lead, Bandcamp), Edward Balassanian (CEO, Aimi), Jen O’Neill (business affairs consultant), Maria Garrido (Deezer), and Marc Azaïs (Director of Business Development and Foresight, SACEM), moderated by Jay Ahern (Chief Strategy Officer, AFEM). The session asks where the lines are and who gets to draw them. Independent operators who have not yet formed a position on AI licensing should treat this as a briefing rather than a discussion.

12:00 - From Data to Decisions: Why Artists and Teams Are Building Their Own Operating System | Summit Stage Outdoor

This session runs in parallel and addresses what has become the central operational challenge for professional artist management and label services: data fragmentation. Artist teams have more data available than at any point in music business history. Streaming numbers, social performance, fan behaviour, marketing attribution. None of it is connected. The managers of Chase and Status and Goldford discuss what centralising that data into a single operating system has unlocked in practice. For independent labels and distributors still managing catalogue data across disconnected spreadsheets and platform dashboards, this conversation is a direct mirror.

12:35 - Beatport Future Vision | Summit Stage Outdoor

Matt Gralen (President and CFO, Beatport) and Helen Sartory (Chief Revenue Officer, Beatport) share what is coming at The Beatport Group in 2026 and beyond. Beatport is the majority stakeholder in IMS and a central distribution and discovery platform for the electronic music market. What they announce here shapes the commercial landscape forindependent labels and distributors across Europe for the next twelve months.

Friday 24 April

The Ninja Tune session closes out the summit on day three. Co-founder Matt Black examines how an artist-led independent ethos has been sustained across a roster that now spans more than seven sub-labels and more than three decades of operation. The structure Ninja Tune built, centred on genuine ownership and creative control rather than distribution dependency, offers a model that many European independent operators are attempting to replicate. The difference is that Ninja Tune built the infrastructure first. Day three concludes with the IMS Dalt Vila closing celebration at Ibiza’s UNESCO World Heritage site.

The Pattern Across Three Days

The through-line across the most relevant sessions at IMS 2026 is not technology. It is the gap between what the independent sector has built culturally and what it has built operationally. Labels with strong catalogues and artist relationships are still running on infrastructure designed for a different era. The royalty distribution systems are incomplete. The data environments are fragmented. The AI licensing frameworks are being written by parties who have no particular reason to prioritise independent interests. The summit is three days of the industry talking openly about these gaps. What happens after the summit is decided by the operators who were paying attention.

Most independent labels and artists approach a release as an event. Months of preparation, a single window of promotion, and then the wait to see how the numbers land. If the numbers disappoint, the instinct is to wait longer next time, refine more, get it right before putting it out. That instinct feels logical. The data says it is wrong.

The platforms that determine whether your music reaches new listeners reward one thing above almost everything else: sustained activity over time. Not perfection. Not a single standout release. Consistent presence, deepening catalog, and accumulating engagement signals. Understanding why changes how you think about release strategy.

How streaming algorithms actually work

Spotify's recommendation system, which drives the bulk of discovery for most independent artists, operates through a handful of core signals. The ones that matter most are not total stream count. They are the stream-to-listener ratio (how many times each unique listener plays a track), save rate (what percentage of listeners save it to their library), skip rate (how quickly listeners abandon a song), and playlist adds. These signals tell the algorithm whether a song is genuinely resonating, not just whether it was heard.

What the algorithm does with those signals is important to understand. It does not make a single decision at release. It runs a continuous test. A new track is served in small batches to a limited audience. If the engagement signals from that test audience are strong, the track is pushed to a larger audience. If the signals are weak, it stops. A song that performs modestly but consistently over six weeks will often outperform a song that spikes in week one and drops off. The algorithm reads sustained engagement as a stronger signal than a burst.

Release Radar, Spotify's weekly new-release playlist, adds another layer. Every time you release a new track, that track is eligible to appear in Release Radar for every user who follows or regularly listens to your music. Note that since late 2024, Release Radar is capped at 30 tracks per user, which means high-volume release weeks can reduce visibility for any single artist. The practical implication is that

spacing releases thoughtfully and building an engaged follower base increases the likelihood of appearing. Release nothing for six months, and you have six months of Release Radar appearances you never made.

What catalog depth does for discovery

The second mechanism is less obvious but equally important. When a new listener discovers one of your songs, whether through Discover Weekly, a playlist, a recommendation, or a social post, the first thing the

algorithm looks for is more material to serve them. An artist with only a handful of releases gives the algorithm very little to work with. An artist with a substantial back catalogue, built consistently over time with strong engagement patterns, gives the algorithm a rich picture of who this artist is and who their audience is. That picture is what enables the recommendation engine to surface your music confidently to new listeners who share the taste profile of your existing fans.

This is the compounding dynamic that matters most for independent labels. Spotify's collaborative filtering system builds listener taste profiles continuously. Every time someone who likes Artist A also saves or replays a song by Artist B, it creates a connection between those two artists in the model. The more material an artist has, the more of those connections accumulate, and the more confidently the algorithm can recommend them to the right new listeners. A deep catalog is not just an archive. It is the dataset the algorithm uses to understand and distribute your music.

Spotify's own Loud and Clear data supports this. Nearly a quarter of the 12,500 artists generating over

$100,000 in royalties in 2024 were not even releasing music professionally five years earlier. Success in streaming is not built on a single moment. It is built on accumulation: more releases, more data, more algorithmic confidence, wider reach.

The perfection trap

The creative instinct to release only when something is ready is understandable. Releasing imperfect work feels like a professional risk. But the streaming economy does not penalise an average track the way a physical release cycle used to. A song that finds its audience is a success regardless of when it was created or how long it took. A song that never comes out contributes nothing to your catalog, your algorithmic profile, or your audience's relationship with your music.

The more common cost of waiting is not a reputational one. It is a data cost. Every month without a release is a month without new engagement signals entering the system, a month without Release Radar appearances, and a month without the algorithm building its model of who your audience is. A six-month gap between releases does not reset the clock. It creates a gap in the data record that takes time to rebuild.

This does not mean quality is irrelevant. Skip rate and save rate are core signals, and a track that listeners abandon quickly will actively work against your algorithmic standing. The argument is not for releasing anything. It is for releasing consistently rather than holding everything back for a moment of maximum readiness that may never arrive, and that the algorithm will not reward more than it rewards sustained presence.

What a consistent release strategy looks like in practice

The cadence that most industry practitioners point to is a new release every six to eight weeks. That is frequent enough to maintain a consistent presence in Release Radar, regular enough to keep accumulating algorithmic signals, and spaced enough that each release gets its own promotional window before the next one arrives.

For labels managing multiple artists, the implication is that release scheduling across the roster matters as much as the quality of individual releases. A label with six artists releasing once per year has six release events. The same label with six artists each releasing every six to eight weeks has roughly forty-five release events per year, each generating algorithmic signals, each appearing in Release Radar feeds, each building catalog depth. The aggregate effect on the label's streaming presence is substantial.

There is also a catalog maintenance dimension. Older releases do not disappear from the algorithm's consideration when new music comes out. A new listener who discovers your latest single and saves it may have tracks from two years ago served to them in their Daily Mix the following week. A deep, consistently performing back catalog keeps generating revenue and audience data long after the promotional window of each individual release has closed.

The metrics that tell you if it is working

Tracking raw stream counts gives you an incomplete picture. The metrics that reflect algorithmic health are different and available in Spotify for Artists.

Stream-to-listener ratio. A ratio above 2.0 means each listener is playing the track more than once on average. Above 3.0 is strong. This is the metric the algorithm uses to identify genuinely engaging music.

Save rate. The percentage of listeners who save the track. Double digits is a healthy benchmark. Saves signal intent and create long-term replay cycles that contribute to sustained algorithmic placement.

Algorithmic playlist reach. The streams breakdown in Spotify for Artists shows what proportion of streams came from algorithmic playlists versus editorial, direct, or external sources. A growing algorithmic share over successive releases is the clearest sign that catalog depth is working in your favour.

Listener geography over time. As catalog depth grows, watch where new listeners come from. The algorithmic spread to international audiences typically accelerates once a certain catalog threshold is passed, which aligns with Spotify's own data showing that more than half of an artist's royalties come from outside their home country after roughly two years of consistent presence.

AI has moved from a side conversation to the central debate in the music industry in under two years. Not because everyone is enthusiastic about it, but because the platforms are already building it in, regulation is catching up, and the decisions you make now will determine where you stand in three years. Here is where things are today: what the platforms are doing, what works in your favour, where it creates friction, and what you as an independent label or artist can do about it.

What the DSPs are doing

The major streaming platforms are each moving in their own way, but the direction is the same: AI is being built into how music is discovered, distributed, and consumed.

Spotify has placed AI at the centre of its listening experience. With Prompted Playlists, users can describe in plain language what they want to hear, and Spotify generates a playlist based on listening history and current music trends. Spotify also rolled out a ChatGPT integration, allowing users to connect their accounts to OpenAI's chatbot for music and podcast recommendations based on mood, genre, or topic. At the same time, Spotify is responding to the downside: it announced changes to its systems for artist verification, song credits, and identity protection, in direct response to bad actors using AI to flood streaming services with content designed to divert royalties away from real artists.

Apple Music introduced Transparency Tags in March 2026, across four categories: artwork, track, composition, and music video. Labels and distributors determine what qualifies as material AI use, and tags are mandatory for new content going forward.

Deezer goes furthest in active detection. The platform tags AI-generated music and excludes it from algorithmic recommendations. Deezer has reported that 39% of daily uploads are AI-generated, and that up to 85% of streams from AI music are fraudulent.

TikTok is the strictest on enforcement. Unlabelled AI content results in immediate strikes, not warnings. In the second half of 2025 alone, 51,618 synthetic media videos were removed.

Bandcamp took a different approach entirely, banning AI music outright in January 2026.

Spotify is also exploring a new revenue stream through AI derivatives. Co-CEO Gustav Soderström described AI remixes and covers as an untapped opportunity for artists to monetise their existing IP, drawing a parallel with how existing IP is monetised in film and television.

What works in your favour

For independent labels and artists, three developments create structural advantage.

Better discovery. Spotify's head of Global Music Curation has stated that Prompted Playlists creates more opportunities for emerging artists, because the tool can go where curated playlists cannot: specific enough that niche artists reach exactly the right listener. AI recommendation systems are becoming smarter, and the more listening data an artist has, the better they surface. For indie artists with a loyal but small fanbase, that is good news: it becomes less about volume and more about fit.

AI as a production tool. AI tools for mastering, stem separation, metadata analysis, and marketing automation are becoming mainstream. AI is already being used for music recommendations, analytics, stem separation, automating marketing tasks, creating visual assets, generating contracts, mixing and mastering, catalogue analysis, and fraud detection. An independent label that uses this strategically has an operational advantage that was previously reserved for larger organisations.

Consent-based licensing. Artists can now participate in verified licensing systems that pay royalties when their style or compositions are used to train AI. That is a new type of revenue stream that did not exist five years ago.

Where it creates friction

The most concrete threat to independent labels and artists is not AI as a creative tool. It is AI as an instrument for royalty dilution.

The pro-rata model of streaming makes all content a participant in the same royalty pool. When

AI-generated tracks are uploaded at scale, every human artist's share shrinks automatically, regardless of whether their fans listen to AI music. Spotify acknowledged this risk explicitly: unchecked behaviour by bad actors can dilute the royalty pool and reduce attention for artists who play by the rules.

Deezer's data makes the scale concrete. When 39% of daily uploads are AI-generated and 85% of those streams are fraudulent, this is no longer a marginal issue. It is an operational threat to trust in the entire system.

The legal situation around training on existing catalogues is also unresolved. While the major labels are striking deals with AI platforms and collecting licensing fees, independent artists and labels largely operate without that protection, meaning their music may be used for AI training without their knowledge or compensation.

Finally, disclosure obligations. Every platform operates under different rules, and the consequences of non-compliance range from takedowns to permanent strikes. DSPs are tightening their policies around AI-generated content. Distributors that build compliance tools help artists navigate this. Those that do not leave their clients exposed.

Where things stand now

The music industry is in a transition. The major platforms accept AI music, but are simultaneously building systems to contain abuse. Legislation around training on existing catalogues is moving but not yet settled. And the balance of power between major labels, which have already secured licensing deals, and the independent sector, which largely has not, is uneven.

2025 was the first year of legitimisation: the lawsuits of 2024 are now transforming into sanctioned partnerships. That means the rules of the next five years are being written now. Those who move with it have a structurally stronger position than those who wait.

 Four things you can act on now

Get your disclosure right. Every release now requires AI disclosure at most distributors. That applies to artwork, audio, and video separately. Labels and artists who do not handle this risk takedowns without warning, particularly on TikTok. Check which fields your distributor requires and complete them, even if no AI was used.

Understand how your metadata and listening data drive discovery. Spotify's Prompted Playlists and AI DJ operate on metadata and listening patterns. An artist with accurate genre tags, complete credits, and an active listening history has a higher chance of being surfaced to the right listener. Metadata is no longer just administration. It is your discoverability.

Build your catalogue actively. AI recommendation systems reward depth. An artist or label with three releases has fewer opportunities than one with fifteen. Consistent output over a longer period gives algorithms more to work with, and gives you more data points to act on.

Monitor your royalty reports monthly, not half-yearly or annually. The combination of AI spam, platform policy shifts, and pro-rata dilution makes the relationship between streams and royalties less predictable than it used to be. Labels and artists who track their reports monthly spot anomalies earlier and can act faster.

The music industry is not changing in ten years. It is changing now, in the platforms you use, the rules being locked in, and the tools becoming available. The independent sector does not have a department for this. But it does have the agility to move faster than a major. That advantage is only worth something if you have the information to act on.

Spotify published its annual Loud & Clear report on March 11, based on 2025 royalty data. It is the most detailed public overview of how money moves through the streaming economy. At Stormi Capital, we work every day with independent labels and artists on exactly the questions this report raises: how to earn more from your music, how to grow internationally, and how to build a career that lasts beyond a single release. Here is what the data says this year.

The pyramid everyone cites, but few actually read

Spotify paid out $11 billion to the music industry in 2025, bringing its lifetime total to $70 billion. Those are large numbers, but the most telling figure sits at the bottom, not the top.

The 100,000th highest-earning artist on the platform generated more than $7,300 in royalties in 2025. Ten years ago, that same position was worth around $350. A twentyfold increase in one decade, while the artist in the tenth spot grew sevenfold over the same period. The floor of the professional music market is rising faster than the ceiling. That is the real story behind the $11 billion figure, and it is relevant for everyone managing a roster or building a career on one.

More than 13,800 artists generated at least $100,000 from Spotify alone in 2025, nearly 1,400 more than the year before. More than 1,500 crossed the $1 million threshold. Artists you may not recognise by name, but who are building full careers. The category of artists running a sustainable music business is larger and more accessible than it has ever been.

Independent labels and artists are half the market

The most strategically relevant number for independent labels does not appear in the headlines: independent artists and labels together account for half of all royalties Spotify pays out. Five billion dollars, from one platform, flowing to the independent sector.

Spotify represents roughly 30% of global recorded music revenue, but for the independent sector that figure rises to more than 50% of streaming income. Indies over-index on Spotify. This is not a coincidence. It reflects a model that rewards catalog depth, niche loyalty, and international reach. Exactly the strengths of a well-run independent label or an artist with a dedicated fanbase.

More than a third of the artists who earned over $10,000 in 2025 were fully independent or started that way. The artist considering a label partnership today arrives with more data, more audience, and a stronger negotiating position than five years ago. That changes how those conversations go, on both sides

of the table.

International growth starts earlier than you expect

On average, more than half of an artist's royalties come from outside their home country after just two years on the platform. Not after a world tour. Not after a sync placement in an American series. After two years of consistent presence.

Artists generating more than $500,000 came from 75 countries, up from 66 the year before. At the $10,000 level, artists represented more than 150 countries. Songs in 16 languages reached the Global Top 50, compared to 8 in 2020. The fastest-growing genres were Brazilian funk (+36%), K-Pop (+31%), Latin trap (+29%), Latin urban (+27%), and reggaeton (+24%). Genres built outside traditional label structures, now leading the financial charts.

For independent labels working with artists in non-English-speaking markets or emerging genres: the data confirms that distributing early and broadly is not a risk. It is the strategy.

Three questions every independent label should be asking

Are you actively using the data available to you? Geographic listening data, demographics, and comparable artists are accessible to everyone who works with them. That information tells you where an artist's audience already exists before you spend budget, which markets are ready for a next step, and which releases are building traction that does not yet show in top-line numbers.

Is your roster set up for international distribution? Consistent presence across multiple markets is not a luxury reserved for large labels. It is the basic infrastructure for any artist with serious ambitions. Labels that organise this early see the return two years later, not five.

Are you thinking in catalog value, not just releases? Spotify generated more than $1.5 billion in concert ticket sales for artists in 2025. A strong catalog builds not only royalty income but audience density that translates into live revenue, merchandise, and long-term value. That is the calculation that matters for labels thinking in multi-year careers.

Loud & Clear is Spotify's own perspective on the industry, and it is worth reading as such. But the underlying data is real, and the direction is clear: the independent music market is growing, going global, and professionalising. The artists and labels that translate this into concrete decisions fastest have a structural advantage.