Suno Filed to Dismiss One Lawsuit – But the Real AI Music Fight Is Much Bigger

When AI music platform Suno filed a motion to dismiss most of the class action lawsuit brought by independent country artist Tony Justice and others, it was easy to treat this case as the defining showdown between AI and musicians.

Entertainment attorney Krystle Delgado – known online as Top Music Attorney – has become one of the most visible faces of that indie artist challenge. In her recent video, she walks through her team’s opposition to Suno’s dismissal bid, positioning it as a high-stakes fight over the future of independent creators and AI.

But here’s the thing most headlines miss: this is only one of several major legal fronts Suno is facing. The big three record labels, rights societies in Europe, and now independent artists are all pressing their own cases in different courts, with slightly different theories and goals.

As someone who uses Suno extensively to create original, branded music and storytelling content, I’m watching all of this closely – not because I agree with every claim being made, but because these cases will shape what AI music creators can and cannot do in the years ahead.

The Indie Class Action Suno Wants Trimmed Down

The case Delgado is talking about in her video is a U.S. class action led by Tony Justice and other independent artists. In broad strokes, they allege that:

• Suno scraped tens of millions of songs – including indie catalogs – without permission.
• Those recordings were allegedly used to train its music-generation models.
• Suno’s outputs can, in some circumstances, get too close to copyrighted works.
• Suno bypassed YouTube’s technical protections (rolling cipher) to mass-download audio for training.

Suno responded with a motion to dismiss most of the claims, arguing that:

• the plaintiffs haven’t shown that their specific songs were infringed in any output, and
• that they shouldn’t be allowed broad discovery without first presenting concrete examples.

Delgado’s opposition brief pushes back on that logic. Her position is straightforward: you can’t expect artists to prove what happened inside a closed AI system before they’ve been allowed to see how it works. That’s what discovery is for.

On a process level, I don’t disagree. The courts should test Suno’s arguments. If there was unlawful scraping, it should be surfaced. If the DMCA’s anti-circumvention rules were broken, that matters.

Where I part ways is with how this indie lawsuit is being framed in relation to the much larger legal picture – and what it implies about AI-assisted music generally.

The Bigger Picture: Suno vs. the Majors, Koda, GEMA and More

The Tony Justice case is not the only, or even the biggest, legal battle Suno is involved in. It’s one node in a much bigger network of disputes:

• Major labels & RIAA (U.S.) – Sony Music, Universal Music Group, and Warner Music, via the RIAA, sued Suno and Udio in 2024 over alleged mass copyright infringement in training data and, more recently, DMCA “stream-ripping” claims tied to YouTube’s rolling cipher.
• Collecting societies (EU) – organizations like Germany’s GEMA and Denmark’s Koda have filed their own cases, arguing that Suno used repertoire under their control without licensing and without paying rightsholders.
• Udio settlements vs. Suno’s ongoing fight – Udio has already begun settling with major labels and moving toward licensed, opt-in AI music platforms, while Suno is still in active litigation and simultaneously raising large funding rounds.

In other words, Delgado’s case is one of the “smaller” suits in terms of industry impact, but it still matters – especially symbolically – because it’s framed as a grassroots push by independent artists rather than the big corporate catalog owners.

My issue isn’t with her right to bring that case. It’s with the assumptions about AI-assisted music that sit behind it and how little attention is being paid to how creators are actually using Suno in 2025.

My Position: Support the Process, Question the Premise

Let me be clear:

• I support the right of any artist or rights holder to sue if they believe their work was misused.
• I support courts examining whether Suno’s data collection and training practices complied with the law.
• I support the idea that AI companies should not get a free pass just because the tech is new.

What I don’t agree with is the implied idea that:

If an AI model was trained on any unauthorized material at any point, then all AI-assisted music made with that tool is inherently tainted, even when you can’t point to a specific “original” it supposedly copies.

As an AI music creator, I:

• write my own lyrics,
• build my own fictional universes and characters,
• shape my own sonic identity,
• use Suno to realize ideas I didn’t have the budget or technical skill to produce otherwise.

If someone is going to call those songs “derivative,” they need to show what they’re supposedly derived from: which melody, which lyric, which track, which performance.

Without that, we’re not talking about enforcement anymore. We’re just casting suspicion on an entire generation of creators because they use modern tools.

A Realistic Look at Suno’s Output (From an Actual Long-Time User)

I’m not here to pretend Suno has always behaved perfectly.

My real-world experience:

• I did see a few tracks flagged for copyright issues in 2024.
• I did encounter the occasional stray producer tag in early versions.
• I did hear of edge cases where the vibe felt a bit too close to something known.

That matters. Those early glitches are part of the story.

But they were:

• rare, not constant,
• usually patched or mitigated quickly,
• far less common by the time V4.5 rolled out,
• and in my current experience with V5, extremely rare when you’re writing your own lyrics and steering your own sound.

If you use modern versions of Suno and actually try to copy a famous song, you’ll generally find it doesn’t cooperate. Style, mood, and vibe are absolutely influenced by your prompts – but direct one-to-one cloning is not how the system behaves in day-to-day use.

So yes: early versions had issues. Yes: those issues deserve scrutiny. But no: that does not mean every AI-assisted track made in 2025 is some uncredited remix of a hidden source.

The Overlooked Frontier: Uploading Your Own Human-Made Songs into Suno

Ironically, the most important legal and ethical question for today’s creators is barely in the conversation:

What happens when a musician uploads their own human-created work into Suno to create derivatives – and how protected is that work inside the model?

This is not about scraped catalogs or old label disputes. It’s about:

• independent artists who own their masters,
• creators who write and record original songs,
• songwriters who upload stems to experiment with style and arrangement,
• AI music creators who want to evolve their own IP, not borrow someone else’s.

These creators are using Suno in a legitimate way: upload your own track, transform your own track, develop your own project. Completely within the spirit of copyright.

The question they have is different:

Once my song is inside Suno, is it locked to my account and my sessions – or can its patterns ever influence what someone else generates later?

To answer that, we’d need clear transparency on things like:

• Does Suno train on user uploads by default, or are they processed in isolation?
• Is there a “do not train” or “session-only” mode for uploads?
• Can patterns from my upload ever bleed into the broader model behavior?
• What happens to all the intermediate versions a creator makes before the final, human-recorded release?

Courts, labels, and law firms are focused heavily on old catalog training. Meanwhile, a new wave of artists is trying to figure out whether it’s safe to push their own best songs into these systems.

Should There Be Explicit Upload Controls?

A more forward-looking discussion would be about product design, not just punishment:

• A “private upload” mode where content is never used for training.
• Clear labeling of whether and how user uploads influence model weights.
• Options to keep AI-assisted derivatives fully siloed to the original creator.

Other AI sectors already lean this way. Some tools give users the option to keep their content out of training by default. There’s no reason AI music platforms couldn’t do something similar – especially for uploaded, fully human-authored works.

That kind of control would do more to protect independent artists using AI today than any abstract argument about AI being inherently “derivative” just because it was trained on large datasets.

AI Tools Don’t Eliminate Human Musicians – They Send Them More Work

One more point that often gets lost in these debates:

For many digital creators, AI tools are not a shortcut around human musicians, producers, or designers – they are a bridge to them.

In my own workflows:

• I use Suno to develop concepts, moods, and story-driven musical ideas.
• I iterate lyrics and structure until the narrative feels right.
• I then take those ideas into BandLab or a DAW and shape them further.
• I look to human collaborators to re-record, refine, and elevate those ideas into final, commercially released works.

That means:

• more commissions for musicians and vocalists,
• more production and mixing gigs,
• more design work for cover art and branding,
• more creative partnerships overall.

AI lowers the barrier to starting a project. Human talent is still what finishes it at the highest level.

So Where Does Suno’s Dismissal Bid Leave Us?

Suno’s attempt to dismiss or narrow the indie class action is just one step in a much larger sequence:

• Big labels and the RIAA are pressing broader claims about training data and DMCA violations.
• European societies are testing similar theories in their own courts.
• Some AI companies (like Udio) are already pivoting toward licensed, opt-in models.
• Others, like Suno, are still very much in the litigation trenches.

I don’t believe Delgado’s case – in its current framing – adds much practical protection for AI music creators who are using these tools responsibly. But I do believe the process has value, because it forces hard questions onto the table:

• How was training data actually acquired?
• Where does fair use stop and infringement begin at scale?
• How should anti-circumvention rules apply to AI training workflows?
• What rights should creators have over music they upload into AI systems?

Those are questions worth answering. I just want to make sure the answers don’t erase or delegitimize the work of AI music creators who are building new, original projects in good faith.

A Question for AI Music Creators and Musicians

I’ll end with a question for you – not for the labels, not for the law firms, but for the people actually making music:

If you upload your own original human-made music into Suno to create derivatives, what kind of control or protection do you believe you should have over how that music is handled inside the platform?

For example:

• Should your upload be completely isolated so it never affects anyone else’s output?
• Should there be a “non-trainable” or “private upload” mode for your songs and stems?
• Should AI-assisted derivatives generated from your uploads be locked to your account only?
• Or do you think shared training on user uploads is acceptable if it’s clearly disclosed?

If you’re an AI music creator, a traditional musician testing these tools, or a producer trying to navigate this new landscape, your real-world perspective matters more than any abstract legal theory.

Share your thoughts. Because no matter how these lawsuits end, it’s the creators who will have to live with the systems we build next.