AI-made songs have climbed real charts. A virtual R&B singer reached a Billboard radio chart, and AI-assisted country acts topped a digital sales chart. Now the platforms and the record labels are drawing lines. For a business that uses music in ads, social videos, on-hold lines or in-store playlists, those lines decide whether a cheap AI track is a smart shortcut or a legal headache.
What happened
In August, Spotify said it will label artist profiles that represent AI-generated identities as "AI Persona" and, by default, keep their music out of editorial, algorithmic and personalized recommendations unless a listener chooses to follow the artist. TechCrunch reports that Spotify will not rely only on artists to disclose this, and will review profiles that appear to use photorealistic AI-generated names and images. TechCrunch on Spotify's AI Persona labels.
On September 9, the AI song generator Suno replaced its models with a new family, v6, which it says was built using licensed data from Warner Music Group, BMG and the distributor Believe, and not the music used to train its earlier models. TechCrunch notes that Sony and Universal were still suing, and that an artist had filed his own suit. TechCrunch on Suno v6.
Nine days later, Variety reported that Sony Music and Universal filed a new suit arguing the licensed model still infringes, because training a new model on the outputs of an older one does not cure the problem. Suno called the claims flawed on the facts and the law. Variety on the new Suno lawsuit.
Billboard keeps a running list of AI and AI-assisted acts that have reached its charts, including R&B singer Xania Monet and the country act Breaking Rust. Billboard's list of AI artists on its charts.
Why it matters for your business
Most owners will never release a song. But plenty use music every day, and AI tools make it tempting to generate a jingle or a background track in a minute for a few dollars.
The legal ground is still moving. The biggest AI music company spent 2026 replacing its models because of lawsuits, and the largest labels still say even the new one infringes. Courts have not settled any of this. If you use AI music commercially, you are relying on a tool's terms of service in a contested area.
Platforms are sorting AI music into a separate lane. Spotify's labels and recommendation rules mean an AI-made brand track is less likely to be discovered there. If your plan was to grow an audience with an AI "artist," the plan just got harder.
Customers notice. A label on the profile, a comment under your video, a reviewer who spots it. Some audiences do not care. Others feel tricked when they learn a "band" is not real.
To be fair to the tools: licensed training is real progress, and AI music is genuinely useful for drafts, mood tests and background audio where nobody expects a human performer. Plenty of stock music was never art either.
What we'd do
- Read the license, not the marketing. Before using any AI track commercially, check that the plan you pay for grants commercial rights, and save a copy of the terms on the day you generate it.
- Prefer licensed sources for anything public. For ads and brand videos, use stock libraries or tools that state their training data is licensed, and keep the receipt.
- Use AI music for drafts first. Rough cuts, pitch videos and internal training are low-risk places to start.
- Do not invent a fake artist. If you use AI music, say so where it matters. Do not present it as a human band.
- Never imitate a real singer. Voice clones of known artists are exactly what the platforms and courts are moving against.
- Check your in-store music service. Ask your provider whether its catalog includes AI tracks and how they are licensed.
Our view: AI music is useful and here to stay, but in 2026 it is still being sorted out in court. For anything a customer will hear, pay for clear rights and be open about what you used. The few dollars you save on a generated jingle are not worth becoming an example in someone else's lawsuit.
Want a system like this in your business?
A discovery conversation is where we map it: what to automate first, where the human checkpoints go, and what it would take.
Book a discovery conversation →