Musixmatch Eyes Larger Role in Music Tech, Inks Deal With Suno for AI Copyright Tool

Musixmatch: ‘Lyrics are becoming the LLM of the music industry’

September 28, 2026 By Tim Ingham

You can still just about get away with calling Musixmatch a “lyrics company” – but doing so might raise a quizzical look from Massimo ‘Max’ Ciociola .

Continue to article...

The Italian entrepreneur acknowledges that his company has built its reputation on cataloging the words behind the songs – at last count, working with almost 250,000 rightsholders, and being used by over 80 million people worldwide (Musixmatch’s lyrics scroll under songs on Spotify , YouTube , Apple Music , Instagram , and many more). In fact, Ciociola agrees that lyrics will probably always form the bedrock of Musixmatch .

Yet by archiving and policing lyrics in over 250 languages, he notes, Musixmatch is already established as a “sophisticated royalty management company” – monitoring usage online, and paying through royalties to rightsholders large and small.

Speaking to MBW in the summer air of New York’s meatpacking district, Ciociola notes: “We have evolved into a kind of fintech company – our main job is processing royalties.”

What comes next? Plugging Musixmatch’s lyrical foundation into the new world of prompt-based content, while accelerating the firm’s industry services by embracing new technologies.

As Ciociola puts it: “When we started, lyrics were broken. They were the most-searched term across any geography on Google , but the value chain didn’t work. It was very complex, everyone was scraping and not paying for rights. But we solved it. And as we solved it, music entered the world of ‘eyes on a screen’ – it was no longer just about listening.

“That has opened up a world of possibilities. The focus now is becoming a multi-agent rights administration company, helping protect rightsholders, and helping songwriters protect their own work.”

> “Music has entered the world of ‘eyes on a screen’ – not just listening.”

On that topic, the past year has been a transformational one for Musixmatch , which was founded 16 years ago by Ciociola in Bologna, Italy.

In October 2025, Musixmatch quietly announced that it had inked “AI innovation” deals with all three major music publishers – securing a catalog of 15 million songs with which to build tools that enabled “responsible AI development”.

At the time, all of that sounded a little theoretical and ‘sandbox-y’. But in 2026, it’s started to bear real fruit – and look impressively ahead of the curve.

First, in March, Musixmatch announced Sentinel , a real-time copyright detection service designed for user-generated content (UGC) platforms. Sentinel, developed following investment from Musixmatch-backer TPG , is able to figure out (as someone is typing a prompt) whether even partial infringement of lyrical copyrights has taken place.

The company intends to expand Sentinel’s capabilities beyond lyrics into a broader menu of copyright identification services. And it’s already bagged a keystone client: Suno . The controversial AI music company confirmed in August that it was adopting Sentinel’s tech to “identify when copyrighted compositions, lyrics, and music may be present.” In practice, when a Suno user tries to create a piece of music by inputting the copyrighted lyrics of an existing song, Sentinel will catch it.

Sentinel will also scan the outputs of both AI music and UGC music more widely – on YouTube, for example – to detect the use of copyrighted lyrical fragments.

One major advantage for rightsholders? A near-instant tool for music publishers and labels to discover (and claim against) cover versions on UGC platforms. Sentinel could hardly have arrived at a more timely moment. Look at regulation: the EU AI Act now presses generative-AI developers to show their models are trained in a way that respects copyright, while US takedown rules push platforms to catch infringing material fast. Sentinel, Musixmatch says, is built for exactly these requirements, returning a verdict in under 100 milliseconds.

Meanwhile, Warner Music was the first major to license Suno last year; BMG followed this summer. And even as Universal Music Group and Sony Music Entertainment press on with their lawsuits against Suno, the platform has publicly committed to watermarking and detection ahead of launching a new, licensed model.

The result is a music-AI market suddenly required to prove what it is – and isn’t – using. That is precisely what Sentinel sells, and it lets Musixmatch supply both camps at once: the rightsholders policing their catalogs, and the AI companies trying to stay legal. Musixmatch is positioned as neutral infrastructure, rather than a participatory combatant.

Rio Caraeff , the founder of Vevo , who joined Musixmatch in 2023 and is now Co-President, says that Sentinel’s Suno deal is indicative of a “broader landscape” that calls for “robust infrastructure to help platforms understand what is being created and whether copyrighted material is being submitted”.

Built alongside Sentinel, Musixmatch has also taken the wraps off Music Lens , dubbed the world’s first AI “music agent”. The tool allows music publishers to analyze meaning, mood, sentiment, and context across the metadata contained in their catalog – even when that catalog spans millions of compositions. This can surface key insights regarding the royalty framework of trending music – but also, for example, can produce AI-informed catalog picks to match a potential licensor’s sync brief.

To Ciociola, the market opportunity ahead of Musixmatch is vast, and powered by two themes: (i) Lyrics becoming an ever-more important – and increasingly monetizable – component of songs; and (ii) AI technology enabling Musixmatch’s engineers to rapidly increase the company’s tool-building productivity.

The under-monetization of lyrics is often spotlighted at the National Music Publishers’ Association ’s AGM. For example, the NMPA estimated that lyrics licensing was responsible for just 1% of publishers’ and songwriters’ revenues in 2024. (More recently, lyrics income has been folded into an ‘Other’ category.) In Ciociola’s telling, that 1% radically undersells the potential market. Lyrics, he says, are the connective tissue between a song and whatever meaning a listener takes from it – and the revenue line badly lags their true strategic worth. As discovery tilts toward AI, he believes that gap will close fast, with the category set to explode in importance – and value.

“In the language of AI, lyrics are becoming the LLM of the music industry,” says Ciociola. “They are literally the language of the music.” Search and discovery of music as we know them, he predicts, will soon “be forever gone,” replaced within a few years by prompts: build me a playlist of songs about heartbreak , rather than a query for a track by name. Lyrics, he suggests, will be at the center of this shift.

As for AI empowering the development of additional tools at Musixmatch, underpinned by the company’s giant catalog, Ciociola says: “When you pair AI coding with the best engineers, you can suddenly transform a company’s vision into working [prototypes] much faster. It’s like Aladdin’s lamp!”

Another accelerant in Musixmatch’s growth has been TPG. The private equity firm’s investment in Musixmatch in 2022, says Ciociola, was less about a cash injection than about entering “a new league” – trading the seed-stage grind for backers who brought a blueprint for scaling a tech company. (TPG’s other music biz-adjacent investments have included Spotify, CAA , and Fender . More recently, it’s taken an anchor stake in OpenAI ’s new enterprise-deployment venture – a bet on precisely the kind of AI adoption Ciociola is trying to drive in his own company.)

> “We need more curiosity. We need to be under review every day.”

Not everything in today’s music industry, however, gives Ciociola cause for excitement. He balefully notes that, when Musixmatch launched in 2010, it did so surrounded by “amazing young talent breaking through in music tech” – naming Daniel Ek (Spotify), Alex Ljung and Eric Wahlforss ( SoundCloud ), and the teams behind Songkick and The Echo Nest as both contemporaries and inspiration.

“I don’t see young talent breaking through with fresh [music-tech] ideas like that today, and that worries me,” he says.

Ciociola frets that music has become too comfortable letting its technology drift. His prescription is simple. “We need more curiosity,” he says. “We need to be under review every single day.”

That includes music publishers, labels, and collecting societies who, he says, should by now be running on modern software – exposing data through APIs, letting rightsholders pull real-time reports instead of waiting months for a statement, and wiring AI “agents” into functions from finance to catalog management.

> “In the language of AI, lyrics are becoming the LLM of the music industry. They are literally the language of the music.”

The building blocks, he points out, are cheap and available in a way they never were before; what is missing is curiosity, not capability. “There are no excuses” for a company not to modernize, he argues – a conviction that reads less like damning commentary on the industry than a description of the products Musixmatch has spent the past year shipping.

What of the litigation from music rightsholders facing Anthropic – specifically regarding the use of lyrics – as well as Musixmatch’s banner new client, Suno? Ciociola’s instinct is plain. “I fully stand on the IP owners’ side,” he says.

“Theoretically, reading about [AI companies] pirating and scraping books, for example, to train their models? It’s wrong. All of these things need to be fixed. And if you’re an IP owner, you’re right to be mad – it’s like entering a famous museum and stealing things for your own gallery. “But I also think back to when Google and YouTube launched, and how angry rightsholders were, including the news business. Most of that tension was solved by technology, and I’m hopeful it will be similar [with music and AI].”

On Anthropic specifically, he expects settlements will be struck with the music rightsholders. “They absolutely have to fix it,” he says, “and they will.”

More generally, Ciociola suggests that in any business where two parties are reliant upon one another for the next growth spike, “litigation is a failing moment”, adding: “It always wastes time and costs a ton of money.”

Ciociola’s own governing rule, learned from his father, is almost monastic: “You need to sleep well at night. That means no blame, no lies, no secrets.” (This may be spoken with a nod to the litigation Musixmatch itself faces from rival LyricFind , which accuses Ciociola’s firm of anticompetitive practices, citing an exclusive licensing agreement with Warner Chappell . Musixmatch says the claim is “without merit”. Ciociola doesn’t add to that statement in our discussion.)

A technologist by background – he built and sold an early mobile business before music, and later served on the European Commission ’s high-level expert group on AI – Ciociola has always approached the music industry the way an engineer approaches a broken system: by asking what isn’t working, and why. In the early days, this led him to conclude that the sector of the industry seeing the very worst in value “leakage” was also its most essential: songwriters.

Ask him what a songwriter is actually up against today, and Ciociola walks the gauntlet: getting a work correctly registered, having its usage collected across dozens of territories and platforms, actually being paid what is owed, then policing the open web for unlicensed copies. Value leaks at every stage, and the writer is often the last to see it. Closing those gaps with data and software is, in his telling, Musixmatch’s central mission. “When there is an amazing painting, we talk about the painter,” he says. “But when there is a song, we so often only talk about the performer – not about the songwrit

_Originally reported by [Music Business Worldwide](https://www.musicbusinessworldwide.com/musixmatch-lyrics-are-becoming-the-llm-of-the-music-industry/)._