How does iMusician evaluate and label tracks that use AI exclusively for technical processing?

Share & Collaborate Music Production
1

Dear iMusician team,

My name is Rafael, and I am an independent musician and music producer. I am in the process of switching distributors and would like to understand in detail the criteria adopted by iMusician regarding hybrid productions and the use of AI tools.

Context and Production Process:

My workflow combines traditional and cutting-edge techniques, using artificial intelligence models in a strict and punctual way, acting as digital signal processing (DSP) and mastering tools — functioning, in practice, as advanced quantizers.

In contrast, 100% of the vocal tracks are entirely human. I specialize in the Tuvan throat singing technique and record the vocals without any type of processing or effect, dispensing even with traditional tuning tools like Auto-Tune.

The Current Problem:

The distributor I am currently under contract with has been automatically and inappropriately flagging my works as “AI-generated.” This automated flagging does not reflect the reality of my production and causes serious damage to my career:

  • Artistic discredit: As an independent artist starting my journey, the label of fully AI-generated content compromises the perception of the audience and curators regarding my work.
  • Distribution restrictions: I am unable to make my albums available on platforms like Bandcamp.
  • Blocking of essential features: I cannot use fundamental monetization and reach tools, such as YouTube Content ID and YouTube Shorts, due to restrictions against content flagged as AI-generated.

Given this situation, I am looking for a new distribution partner and would like to confirm: how does iMusician evaluate and label tracks that use AI exclusively for technical processing (DSP/mastering), while maintaining purely human vocal performances and arrangements?

Thank you in advance for your attention, and I look forward to your guidance.

Sincerely,

Rafael M. Elias (Hellthoven)

IPI: I0076018772 // CAE: 1343163183 / IPN: 11643350

2

Hey @Hellthoven,

We can provide some basic tagging (“no AI”, “AI-assisted”, “AI-generated”) to specific metadata but ultimately the DSPs have the last word when it comes to flagging content on their own platforms.

Despite the data we deliver, they still use their own detection & flagging tools. If the amount of AI you’ve used is high enough to trigger a tag, then unfortunately the specific shop’s tagging policy will overrule our own.

We understand that you may feel these practices are unfair towards your creative process, but unfortunately it is up to the platforms hosting your music to decide whether or not your music is eligible to specific features.

Best,

3

Hi Carlos,

Thank you for your helpful and prompt response.

However, I have to disagree with you. I’ve run several tests and can prove that AI detection tools are flawed. Furthermore, the stores have stated that labeling is 100% the distributor’s responsibility. I can cite Deezer itself as an example:

“Finally, since your music was delivered to Deezer via a third-party distributor, any further inquiries or requests related to labeling should be directed to your distributor. They can liaise with us directly for additional support.”

This makes things difficult, because it turns into a “passing the buck” situation, as we say here in Brazil, everyone pushing responsibility onto each other, no one taking ownership, and no one clarifying questions concisely.

If you could explain this in more technical terms, I’d appreciate it, because the methods you described weren’t clear.

Best,

4

Hi @Hellthoven,

We understand that you may feel these practices are unfair towards your creative process, but unfortunately it is up to the platforms hosting your music to decide whether or not to tag your music according to their internal tagging policies.

Best,