demorphus
AI detection can matter after a distributor accepts a track: platform policies may affect labels and discovery. To evaluate a release, keep four outcomes separate—availability, classification, recommendations and payment. An improvement in one is not proof of an improvement in all four.
Four outcomes to record for each release
| Outcome | Useful evidence |
|---|---|
| Publication | Release identifier and working public track link |
| Classification | Named detector report or a dated visible platform label |
| Discovery | Audience-source information available in the artist account |
| Payment | Distributor statement tied to a period and recording |
This is a proposed tracking method, not a claim that every platform exposes each field. Mark information as unavailable when you cannot observe it.
Spotify and Deezer require different explanations
Deezer confirms recommendation exclusions for detected AI music. Spotify describes a multi-signal assessment of spam and low-value content. Its AI Persona restrictions also concern artist identity.
The practical implication is to investigate the actual platform and outcome. Read the detailed Spotify recommendations guide or Deezer monetization guide rather than applying one generic “AI penalty” explanation to everything.
Measure the part of the workflow you can control
Before delivery, preserve your original WAV, listen to the final export and establish a detector baseline. If you use Cleaner, compare the returned audio and report. Name the engine and date rather than calling the output universally undetectable.
SH Labs describes its results as probabilistic predictions. Demorphus uses this external measurement to help compare files; the result does not expose platform recommendation systems. Compare Demorphus plans if you need processing, or start with the free checker for an existing file.
What our published release observation adds
We have uploaded thousands of processed tracks to platforms without an AI flag observed. That gives us release experience beyond a standalone classifier test. The results page records this separately from the historical 2,500+ track detector cohort.
The observation does not include a published platform breakdown or a controlled comparison of untreated and treated releases. We therefore do not turn it into a platform-specific success percentage or a claim of additional recommendations and revenue.
Use a consistent follow-up record
- Keep the final audio and pre-release analysis together.
- Record the platform links once the release is live.
- At each review, date any visible label observation and note the view checked.
- Save the audience data available for the same track and period.
- Match revenue to the distributor’s statement when it arrives.
For your next release, begin with how to monetize Suno songs. Use the earnings guide for scenarios and the Spotify workflow for delivery and payment.
Platform sources reviewed October 30, 2026. Check current policies before your release.
Prepared with AI assistance. Sources below.