If you have released music anywhere online, it may already have been swept into a dataset used to train an AI model — often without anyone asking you first. For a lot of independent artists, that is an uncomfortable realization: your work, used to teach a system that might one day generate music competing with yours, with no record of whether you ever agreed to it.
The law here is still catching up. But there is one thing you can do right now that costs nothing and puts your position on record: declare, clearly and with a date attached, whether your music may be used for AI training.
Why just having copyright is not the whole answer
Under the Berne Convention, you own the copyright to your music automatically the moment you create it — no registration required. That is real, and it matters.
But copyright ownership and a documented, dated statement of your terms are two different things. Copyright says the work is yours. It does not, on its own, create a clear record of what you did and did not permit — the kind of record that is useful when a dataset, a platform, or a company later asks whether this artist allowed this.
Right now, most music carries no such record at all. Which means the default, in practice, is ambiguity — and ambiguity tends to favor whoever is doing the scraping, not the artist.
What a clear declaration actually does
A declaration does not build a technical wall around your file. No metadata field can physically stop a determined scraper — anyone who tells you otherwise is overpromising.
What a declaration does is remove the ambiguity. It creates a dated, verifiable statement that says, in your own words: the creator does not permit this work to be used for AI training. That statement matters in a few concrete ways:
- It is on record before any dispute. A position stated in advance — before there is any argument — carries more weight than one asserted after the fact.
- It is an explicit, stated position. As the EU AI Act and similar rules push toward transparency about what goes into training data, a clear opt-out from the creator is the kind of record regulation is starting to ask for. To be precise: this is a declaration recorded on your certificate, not a technical signal embedded in your audio file — no such standard is widely honoured yet.
- It shifts the burden. If your terms are documented and dated, anyone who uses the work for training anyway is doing so against a stated position — not into a vacuum.
It is the difference between nobody knowing what the artist wanted, and the artist having said no, on this date, verifiably.
How to make the declaration
This is exactly what the AI-training declaration on an Audiverify certificate is for. When you document a track, you can state your terms for AI training use:
- Not allowed — the default. If you say nothing, your certificate records that AI training use is not permitted. The restrictive option is the automatic one, on purpose — permission should be something you actively give, never something assumed.
- Allowed — if you do want your work available for training (some creators do, especially for open datasets), you can say so explicitly.
- Allowed with attribution — permitted, but on the condition your authorship is credited.
Whatever you choose is captured word-for-word, stamped with an independent trusted timestamp, and shown on your certificate as your declared position. Audiverify documents that declaration — it does not monitor or control how third parties use your file. The value is the clear, dated record, not a promise to police the entire internet on your behalf.
The honest bottom line
Declaring your terms will not stop a bad actor who is determined to ignore them. Nothing at the file level can. What it does is make your position unambiguous and verifiable — which is the part that has actually been missing.
For years, the practical default for independent music has been silence: no stated terms, no dated record, nothing to point to. That silence quietly favors whoever wants to use your work. A clear declaration flips that — it costs you nothing, and it means that if the question ever comes up, the answer is already on record, in your words, with a date attached.
You made the work. You should get to say what happens to it. This is one small, concrete way to put that on the record.
Audiverify issues timestamped documentation for music releases, including AI-use and AI-training declarations, so independent creators have a clear, dated record of their stated terms. It documents what you declare — it does not grant rights or control how others use your file.
Related reading
Audiverify
Cryptographic fingerprinting, AI disclosure documentation, and dispute-ready evidence workflows for professional music releases.