Anyone who is a keen devotee of the music business’s sultan of sync, the big dog of catchy pop and top banana of the twelve inch, better known as Lucian Grainge, will have been more than a little worried recently. On Universal Music earnings calls over the past few quarters, Papa Grainge has seemed flat, fractious, tetchy and frankly out of sorts.
The trademark pizazz was missing, his garrulous zing and zest gone entirely. Analysts, whispering amongst themselves as they huddle round their Bloomberg terminals, asked where the ebullient Lucian of old has gone.
And we mere mortals on the sidelines listened in, aghast at the listless patter of the big dog himself, knocked for six by a plummeting share price. But as a wise man once said - it’s not the fall that kills you, it’s the sudden stop.
Good news yesterday then, as Universal held its Q4 2025 and full year earnings call. The pizazz is back, the zest replaced. True enough, there was only one single barked “win-win!”, but the lesser-spotted superfan is back in the sights of Papa’s twelve bore, and he’s got a new hobby horse to ride to the moon and back.
Superfans are still the future, AI is the icing on the cake and the AI-powered superfan revolution is going to breathe new life into Streaming 2.0. Hallelujah!
And Lucian… Well, Lucian is - in his very own words - “the hostess with the mostess”, and as far as AI goes he’ll be at “every single dinner party that there is around town”. He’s Ethel Merman singing Irving Berlin and he’ll flash his ankles for any good looking AI who slips him the wink.
It’s been a drama-filled few weeks for AI and the music business. Spotify’s new co-CEO Gustav Söderström held analysts spellbound as he recounted his torrid ‘Heated Rivalry’ Christmas-with-Claude chatbot love affair before boasting that his best engineers don’t waste their time writing code any more.
Over at Warner Music, Robert Kyncl published a shareholder memo betting the company on AI as an “accelerant” - which is technically something you use to start a warehouse fire.
And Jeremy Sirota, who spent six years as CEO of Merlin telling anyone who’d listen that “independent music is not raw material for tech companies to exploit without consent”, realised he’d got it wrong all along and signed up as Suno’s new Chief Commercial Officer - the company and scraper-in-chief that built its $2.45 billion valuation by exploiting pretty much all the music ever created without a whole load of consent from anyone.
And now it’s Lucian’s turn. More on Universal’s big AI play shortly. But first, the main event - the thing we’re all here for. How many gazillions did Universal rake in last year?
On paper, the results are solid. Revenue up 8.7% year-over-year in constant currency to €12.5 billion. Adjusted EBITDA up 8.6% to €2.8 billion. Subscription revenue up 8.6%, six consecutive quarters above 8%.
Matt Ellis, Universal’s CFO, rattled us through the numbers with the clinical precision and mechanical charm of the speaking clock - “another excellent year”, “healthy growth on both the top and bottom line”, “on schedule with our cost savings programme”. Lovely. But everyone knows the real razzle dazzle is elsewhere.
And the creative success stories keep coming. Kendrick Lamar was “the night’s biggest winner” at this year’s Grammys with five awards including Record Of The Year, making him “the Grammys’ most decorated rap artist of all time”.
Olivia Dean was named Best New Artist - “the fourth time in the past five years that a UMG artist has received that honour” - before sweeping four awards at the BRITs including Artist Of The Year. She’s now the bookmakers’ favourite to sing the theme for the next Bond movie, if Amazon and Barbara Broccoli can ever sort out their differences.
Taylor Swift topped the annual IFPI Global Artist Chart. Nine of the top ten were Universal artists - for the third consecutive year.
Lucian, naturally, is covered in glory and taking full credit, flashing his win-win, his ankles and his Hollywood smile, telling analysts he’s “never been more energised”. He’s having the time of his life. He’s never felt this way before. The pizazz is back and the only damper on things is Universal’s plunging share price.
Universal Music Group listed on Euronext Amsterdam in September 2021 at a reference price of €18.50. At its peak in May 2024 it hit €29.49. Yesterday, before the earnings call, the stock closed at €19.36. Today as US markets opened it hit an all-time low of €17.83, a drop of more than 9%. Ouch.
Four and a half years after listing, the stock is a full fifty cents below where it started. Grainge’s megabonus stock options - granted in April 2023 with a strike price of €19.81 - were split into three tranches, unlocked at share price hurdles of €26.50, €30, and €38.
The first tranche was triggered when the stock hit its peak last year, netting Grainge around €15.3 million in additional options over shares. But with the stock now almost €2 below the exercise price, those options are worth nothing: they give him the right to buy at a price above what the shares are currently worth. His other two triggers of €30 and €38 seem impossibly distant.
Grainge made a big bet on his Streaming 2.0 hypothesis - first unveiled in autumn 2024 - and it looks now as though the markets have stopped believing the hype.
Also announced yesterday was the news that the Universal board has shelved a potential US stock market listing - something they were forced into by hedge fund manager, MAGA cheerleader and UMG shareholder Bill Ackman.
The official reason for bailing on the US listing is “uncertainty in the market creating meaningful dislocation in valuations” - which is finance bro for “our share price is horrible, we don’t want to do anything that could make it even worse”, but with Ackman’s current profile, there’s doubtless a quiet sigh of relief at Universal HQ.
The numbers say that Universal’s business is growing - and so the share price ought to be growing too. But the market is not buying what Lucian is selling.
So what exactly is Lucian trying to sell this time around? Well, the superfan is back, and this time it’s brought its buddies.
For three years the UMG superfan and super-premium Streaming 2.0 pitch was top-down. DSPs would build ‘super-premium’ tiers. Fans would pay more. Lossless audio, exclusive content, early access. Win-wins abound and cigars all round.
The only problem was that nobody built that super premium offering. Not Spotify, not Apple, not YouTube, not Amazon. Not one major Western streaming service has launched a superfan offering or super-premium tier.
Spotify’s long-rumoured Music Pro has been delayed so many times that, by late 2024, former CEO Daniel Ek was telling analysts that existing subscriptions were “plenty enough for us”.
The new pitch is bottom-up. Look at these exciting platforms! Look at the ecosystem! Feel the innovation! Grainge namechecked two superfan-specific offerings in his prepared remarks.
The first is Stationhead, a live listening platform that connects artists and fans through “real-time listening experiences, community interaction and integrated commerce”. With over 250 Universal artist events in 2025, Stationhead “contributed billions of premium UMG artist streams” and contributed to “eleven number one albums across the entire industry”.
The second is Even, which provides fans with early access to music, exclusive content and community features. J. Cole used Even for multiple campaigns including the pre-release strategy for his latest album, which “reached hundreds of thousands of fans and sold millions of dollars of physical product”.
It’s so exciting. It’s so innovative. It feels like the glory days of 2011 all over again, when record executives got their expense accounts back, vinyl was definitely fully reviving, and a social listening platform called Turntable.fm was going to change everything.
Anyone who was around the music business thirteen years ago will remember Turntable.fm - the social listening platform where fans DJed for each other in virtual rooms, voted on tracks, chatted and built communities around shared music taste.
Billboard named it top music startup of 2011. It had 600,000 users streaming a million songs a day. Lady Gaga invested. Kanye invested. It was, by any measure, exactly the kind of fan-driven, community-powered, socially connected music experience that Grainge is now breathlessly describing as the future.
The major labels - Universal included - signed licensing deals with Turntable.fm in March 2012. The cost of those deals, and the restriction to US-only users that came with them, killed the service within two years. The founder said the music industry was “an incredibly hard industry to work with” and that the company spent more than a quarter of its cash on lawyers and royalties. It shut down in December 2013.
If that sounds familiar, it’s because we’ve been here even more recently. Clubhouse rode the COVID wave to a $4 billion valuation in early 2021, with every music exec in town hosting rooms and declaring it the future of artist-fan engagement. By 2022 it was a ghost town. The audio social future lasted about eighteen months.
Twelve years on from Turntable.fm and how the turntable has turned, and Grainge, straight-faced, stands up and tells analysts that Stationhead’s listening parties are evidence of the superfan opportunity. It is, functionally, the same product his own company helped destroy.
But here’s the thing that matters more than the irony. Look at what Stationhead actually does. It drives streams on existing subscription platforms. Those streams are already captured in Universal’s subscription revenue line. That’s not new money.
And Even? It’s a white-label storefront that sells physical product - vinyl, CDs, merch. The “millions of dollars of physical product” from J. Cole’s campaign shows up in the physical revenue line, which grew 11.4% in constant currency this year. That’s also not really new money. It’s existing revenue being routed through a different door and relabelled “superfan”.
Grainge also talked up Universal’s own D2C business - “1600 online stores generating hundreds of millions of dollars in revenue” - which is impressive until you remember that “1600 online stores” really only means that someone at Universal found out how to build Shopify stores.
In his annual memo to staff back in January, Grainge highlighted the major’s new retail stores in Tokyo, Madrid, New York and London. All lovely cities, all excellent places for a long weekend, and if Lucian hasn’t got his own private jet to zap him across the oceans he can doubtless borrow Taylor’s - she’s been lending hers out to friends, apparently.
Whether four shops in four capital cities constitutes a superfan revolution is a question the memo did not address.
Which brings us - tediously, inevitably, and with all the weary predictability of a Lucian Grainge annual memo - to AI. We are all, by now, pretty much sick of it. But Lucian, fashionably late as ever, would like everyone to know his thoughts. And boy, does he have a couple.Yesterday’s big pivot - and surprise to no-one who has been following Spotify’s narrative, or read Rob Kyncl’s memo published the day before UMG’s earnings call - is that the superfan delivery mechanism has been quietly swapped from DSP premium tiers to AI-powered interactivity.
Grainge, never one to miss an opportunity to jump aboard a passing bandwagon, has made the same leap - and we can all forget those pesky DSPs and embrace our AI overlords. AI is the future! AI will save music! Look at how AI will deliver value!
Grainge’s AI pitch has two halves that don’t quite fit together. The first half is defensive. Michael Nash, Universal’s Chief Digital Officer, was wheeled out to reassure analysts that AI is not, in fact, eating the business.
Nash came to this knife fight armed with an exhaustively footnoted PowerPoint that made it clear that, actually, nobody cares about AI, and he’s got the research to prove it. Carefully curated focus groups, consumer research with a budget to make a mid-tier pharma company blanch. 28,000 respondents across thirteen countries. Slides. Data points. Statistics oozing from every orifice.
Nash presented a survey which found, among other things, that use of AI is “fast becoming mainstream” with 54% of global consumers expressing familiarity. Among those users, nearly half report conducting music-based queries - what to listen to, what merch is available, what concerts are near me.
Which feels awfully like a Google search, not AI as most people would understand it. But for boomer label execs whose dying BlackBerries were pried from their hands sometime around 2015, ‘the Google’ and ‘the Grok’ are presumably pretty much one and the same.
But when Nash got to the findings that matter, the picture was more encouraging - at least for anyone who - unlike, say, Rob Kyncl or Jeremy Sirota - stands on the human side of the humans-versus-robots fight.
The vast majority of consumers, Nash says, “continue to prioritise human artistry”, with an encouraging 79% saying it is essential to music, with just 4% disagreeing. “76% of music streamers want clear AI-generated labelling”, he continued, and “by an almost 7 to 1 ratio, consumers express disinterest versus interest in so-called AI artists”, while “67% want the ability to filter out AI-generated music entirely”.
In the US, where AI awareness is highest, that figure rises from 67% - two thirds - to nearly three quarters. More than 70% of Universal’s research respondents “insist on transparency, deepfake protection, licensing of training sets”.
All of that is genuinely good news for the music industry - if the research holds up. Nash did not share the survey methodology, the question framing, or how “AI-generated music” was defined for respondents. We’ve asked Universal for more details, and will update separately if we hear back.
Nash also presented consumption data showing that the top ten chart-debuting AI acts - “identified by Billboard and Luminate” - barely register.
“Consumption of this top ten has been immaterial”, Nash said. “The most streamed act didn’t break into the top 7000 globally in 2025, and the number ten act didn’t break into the top 92,000”. In total, the AI top ten account for less than 0.015% of total streams for the top 50,000 artists in the US.
If you’re still standing after that big old bowl of numbers soup, here’s another spoonful: 60,000 AI tracks are being uploaded to DSPs every day, but “85% of AI streams” on Deezer “were identified as fraud and then excluded from royalty pool allocation”. Apple excluded two billion fraudulent streams last year. Spotify “outright removed tens of millions of spamming AI tracks” from its service.
“I’m very aware that a large swathe of the investment community looks at the intersection of AI and media and sees only some of the risks”, Grainge told analysts. “I want to be very clear. We fundamentally disagree with that view”.
Nash backed him up with data showing that Universal’s deals with DSPs include anti-dilution provisions that remove pure AI-generated content from royalty pool calculations, and that “we’re seeing no indication that AI royalty dilution is a material issue for UMG from a revenue perspective”.
Relax, everybody. The zombies are not coming. AI is not going to eat the music business. Which is, as any horror movie fan knows, exactly what someone says five minutes before the rampaging undead come crashing through the windows.
But then - minutes later, same call, same research - Nash pulled a good old switcheroo. “Roughly half of consumers under 45 express interest in AI for music”.
Interest in what? “Deeper personalisation, customisation, remixing, reinterpreting favourite songs, interactive and co-creative music experiences”. So consumers reject AI music by seven to one - but dress it up as “personalisation and remixing” instead, and they love it. Same pig, different lipstick.
And Nash’s slides give even more insight, including an elephant in the room that he didn’t address. That interest figure comes not from the full 28,000 respondents, but from a subset of 8,988 already filtered for being “open to AI in music”. If you survey a bunch of people who say they are open to AI and ask them if they’re interested in AI, you’d be more surprised if they didn’t say yes.
It’s not a finding, it’s a self-selecting foregone conclusion, a tautological contortion of the data that twists it beyond any credible meaning.
It’s also the same distinction Rob Kyncl drew in his Warner memo, where fans would “reimagine” music rather than just listen to it. The majors may not agree on much, but they’ve landed on the same talking point.
For a man whose core message is that AI isn’t a threat and it barely registers, Nash spent an awfully long time explaining why no-one is interested in AI.
Which is a little bit awkward, because the second half of Universal’s big AI pitch is the exact opposite.
Far from no-one caring about AI, it is, says Grainge, “an unprecedented commercial opportunity” and he has “never been more energised about the possibilities”. He listed partnerships with Udio, Stability AI, Klay Vision, Splice and NVIDIA.
Klay’s large music model “is trained entirely on licensed music” and “will evolve AI experiences for superfans”. The Splice collaboration is “building a roadmap for the development of commercial AI tools”. The NVIDIA alliance “will transform and enrich the music experience for billions of music fans around the world”. Grainge quoted NVIDIA as saying that a music catalogue can be “exploited like an ‘intelligent universe’: conversational, contextual and genuinely interactive”.
If that language sounds familiar, it should - it’s strikingly close to what Spotify’s Söderström has been describing with his language-to-music dataset, where natural language becomes the interface and AI becomes the layer between the listener and the music. Same vision, different end of the pipe. Spotify wants to own the interface. Universal wants to own the catalogue that feeds it. Both are describing a world where you talk to AI and music comes out.
Nash went further, calling AI “a paradigmatic change in the landscape with respect to innovation and the evolution of music” - with ‘paradigmatic’ a word it took him three runs to land correctly, in a sentence so dense with empty corporate bullshit that it reads like something produced by an intern at 4am, fuelled by espresso, desperation and Söderström-level-intensity sessions of frantic hallucination-ridden redrafts with Grok.
Not one of these partnerships has launched a product. Not one has a public launch date. Not one came with revenue guidance. The only confirmed AI cash in Universal’s entire results was a single “compensatory payment as part of a strategic licensing agreement with an AI music platform” buried in the Q4 license revenue line. Ellis didn’t name the platform or the amount.
So which is it, Lucian? Is AI immaterial slop that consumers reject seven to one? Or is it a paradigmatic change that will transform the music experience for billions? It can’t comfortably be both at the same time, and yet that is precisely what Universal asked analysts to believe on Thursday evening.
Today’s share price performance suggests that they didn’t buy it either. Guggenheim analyst Michael Morris put his finger on it, asking directly: “Do you expect the majority of that engagement with AI tools to come from new players, or do you expect launches from your DSP partners, and do they have the rights to launch products at this point?”
Grainge jumped in before Nash could respond. The DSPs “feel like our established business partners”, he said, and “what I’ve seen, I’m extremely encouraged by”. He referenced NVIDIA and hinted at “an array of other conversations” with companies “equally as innovative and exciting and well-funded”, before channeling Ethel Merman and Irving Berlin. “We want to be the hostess with the mostest. We want to be at every single dinner party that there is around town.”
But he did not answer the question about whether DSPs currently have the rights to launch AI products - and Nash didn’t come back to it either.
What “hostess with the mostest” actually means, stripped of the 1950s cocktail party charm, is: we don’t control the platform layer, we don’t know who’s going to win, and our strategy is to say yes to everyone and hope for the best. Universal will be at every dinner party - but what’s not clear is whether Lucian will be hosting, or whether he’ll be singing for his supper.
Remember Universal’s big promise from Capital Markets Day in 2024, that a staggering 20% of streaming subscribers would apparently pay double for a “significantly improved offer”? However implausible that sounded at the time, when it was coupled to lossless audio, exclusive content and early access, it at least had the virtue of being based on something people recognised.
That 20% is now back for a second bite of the cherry.
Gone are the hi-fi streams and the backstage passes. Instead it’s AI that people are gagging for. “AI innovation has kind of overtaken the conversation around technology innovation with all the service providers”, Nash said. Hard to say anything otherwise when Spotify’s co-CEOs spent the majority of their own earnings call ensorcelling analysts with their vision of an AI future where - maybe, just maybe - those pesky rightsholders hold less power than they do today.
“We’re going to see AI being a significant component of what will become the super premium tiers of 2026 and beyond”, Nash continued. Whether the 20% who said they’d pay double for better audio are the same 20% who’d pay double for “deeper personalisation and co-creative music experiences” is a question nobody on the call thought to ask.
Nash says the super premium tiers will arrive “in 2026 and beyond”, before adding moments later that AI “could potentially lead to significant opportunity for customer value realisation at the end of this decade and into the next”. So that’s 2029, 2030 or maybe even later - for something that was first promised in 2023.
By the end of the decade that will be seven years of promises, pivots and PowerPoints without a single product to show for it. For context, Instagram went from zero to two billion users in twelve years. TikTok hit a billion in five years and rewired how an entire generation consumes media. Suno - the AI music platform that Universal is still suing - has gone from zero to 100 million users, two million paying subscribers and $300 million in annual revenue in just two years.
Entire platforms rise, reshape culture and generate billions in the time it’s taking Universal and its DSP partners to agree on what a super-premium tier might look like, let alone build one. The major labels are betting their AI future on a landscape where today’s DSPs are still the dominant platforms at the end of the decade. The risk is that this is dinosaurs staring at the asteroid and reassuring each other it won’t block out the sun for long.
You can keep promising that something is just around the corner for a while. Every earnings call is a fresh start, every annual memo a clean page, every Capital Markets Day a chance to reset the clock. The superfan was coming in 2023. Then it was 2024. Then the DSPs needed more time. Now it’s AI, and AI needs until the end of the decade. The narrative keeps moving but the destination never arrives. And the share price keeps falling. But it’s not the fall that kills you.
Universal is being stalked by an activist investor who - depending on which rumour you believe - may want to carve it up and sell it for parts. The US listing that was supposed to unlock value has been shelved.
The headline net profit dropped 26.5% this year - not because the business is shrinking, but because Universal books paper gains on its shares in Spotify and Tencent Music, and when Spotify’s share price came back to earth, an €880 million hole opened up in the numbers. The world’s biggest music company, and its bottom line still moves to the rhythm of Daniel Ek’s share price.
Grainge staked his future - and his megabonus - on a superfan dream that Nash now says is years from fruition. The share price is in the bin and the €30 and €38 per share triggers that unlock Lucian’s next big pay days might as well be on another planet.
And the timeline Nash signalled today - “the end of this decade and into the next” - extends well beyond Grainge’s current tenure as CEO.
Media - including the LA Times and the FT - have already started positioning John Janick, who heads up Universal’s Interscope division, as his natural successor. And you’d better believe they don’t do that without the blessing of Papa Grainge himself. The broken promises - the superfan future that is always just tantalisingly out of reach - may all end up being the next guy’s problem.
But by the time the next guy comes along there’s a real possibility that the cosy relationships between major labels and DSPs that the old guard of music thinks will save them will be gone forever. For all the breathless boasts of being ahead of the curve, having a finger on the pulse of innovation, and all the pivot pivot pivot hustle, the record industry’s upper echelons don’t have the best track record when it comes to spotting the epoch-changing waves before they crash over their heads.
Downloads nearly killed the music business. Streaming came along just in the nick of time, saved it, turned music into an asset class, and locked in unimaginable wealth for the elite executives who - for a decade or so - thought the future was a never-ending golden road.
Lucian, has “seen this, I’ve done it, we’ve managed these transformations. If you really want to go down memory lane, I’ve gone through from LP vinyl into the CD, then into digital downloads. I like what’s going on, I like what I see, and we’re attacking it, and we’re excited by it”.
So it’s natural that there’s an expectation that the same thing will happen again. AI will slot in neatly, the money will keep flowing, and the people in charge will stay in charge.
Grainge told analysts today that he’s “never been more energised”. The pizazz is back. The win-win is back. The superfan is back. AI is the future, and Lucian will be at every dinner party in town. But it’s not the fall that kills you. It’s the sudden stop.
And somewhere between the paradigmatic changes and the hostess with the mostest, between the Stationheads and the Shopify stores and the four lovely retail outlets in four lovely cities, between the seven-to-one rejection ratios and the unprecedented commercial opportunities - somewhere in all of that, the music that Grainge is singing his song to is going to stop.
And when it does, someone will be left standing on their own.
