Learn
Practical guides on AI disclosure, provenance, and proving how music was made — written for people releasing records, not for search engines.
My song was flagged as AI-generated. What do I do?
What an AI flag on your track actually means, the steps that resolve one, why detectors produce false positives, and what evidence is worth keeping before it happens again.
2026-08-06 · 6 min read
AI disclosure when you release music: what you actually have to declare
Where current AI credits appear, why role-level disclosure is more useful than a whole-track label, and how to handle grey areas like mixing assistants and pitch correction.
2026-08-06 · 4 min read
How to prove you made your music — and what actually counts as evidence
No single artifact settles it. What works is a contemporaneous record from several independent sources — what each kind of evidence is worth, and why timing beats volume.
2026-08-06 · 5 min read
How AI music detection actually works, and where it breaks
What detectors measure, why results shift across datasets and audio transformations, how watermarking differs from detection, and why a clean score proves as little as a bad one.
2026-08-06 · 4 min read
Content Credentials for audio: what C2PA does, and what it does not
How C2PA credentials work on audio files, what a manifest actually asserts, why the trust list matters more than the signature, and where audio still lags images.
2026-08-06 · 4 min read
Evaluating AI claims at scale: a policy stack for platforms and distributors
Three independent signals — declarations, detection, and provenance — with what each can and cannot support, and how to combine them without building an appeals queue you cannot staff.
2026-08-06 · 4 min read