FAQ
Everything you need to know about how Ethical Audio works.
Do you train AI models on my music?
No. Ethical Audio is a custom-built AI system — it is not made from any Machine Learning approach, and we do not use ML models of any kind. Everything you upload is processed in isolation, using explicit reasoning and established audio engineering principles, never learned patterns from your music. Your uploads are yours alone.
Is my audio stored?
Not beyond what's needed to generate your results, and it's never used for training or shared with third parties. For analysis, your uploaded audio is deleted from our servers once processing completes — the copy your player uses afterward lives only in your browser, not with us. For mixing, your uploaded stems are deleted once processing completes, and the resulting mix candidates only remain available while you're viewing your results — once you leave, they're gone. We don't currently offer a way to recall or re-download a previously generated mix. We're carefully thinking through the most respectful way to store and manage audio on our end going forward, in a way that gives you full transparency and confidence — but that's planned for future development, not available today. (Mastering isn't live yet, so its data-handling specifics will be published when it ships.)
Can Ethical Audio make changes automatically?
It depends on the service. Analysis never touches or modifies your audio in any way — it only listens and reports back what it finds; your file leaves our systems exactly as you uploaded it. Mixing (and, in the future, mastering) is different: the AI does generate real, processed audio, combining and adjusting your stems into full mix candidates based on what it determines the mix needs — but only after you upload your stems and start a run, and you always choose which candidate, if any, to keep. Nothing is applied automatically in the background without you initiating it.
Who owns the results?
You do. Your music, your decisions, and your releases remain entirely yours. Ethical Audio does not claim ownership or create derivative works without your explicit action.
Is this generative AI?
No. Ethical Audio is assistive AI, not generative AI. For analysis, it never creates anything — it observes and explains. For mixing, and mastering in the future, it does produce real audio, but only by combining and adjusting the stems you already gave it toward a target you chose — it never invents new sounds, instruments, or performances. Every result is guided by your own input, not generated independently of it.
Can I delete my data?
You can delete your account at any time from your profile settings, which immediately revokes access and releases your email for future use. We don't yet offer self-serve deletion of individual tracks or analysis results — reach out at support@eaudio.app and we'll take care of it directly.
How do you ensure transparency?
Our AI shows all of its processing — there are no hidden steps and nothing happens behind the scenes. Every analysis, recommendation, and mix candidate comes with a full, human-readable explanation of how it was reached: what was measured, what references were used, and why. Everything is explainable, from the first metric to the final result.
Who is Ethical Audio for?
Anyone who creates, mixes, or masters music and wants to understand their audio better. From beginners to professionals, our tools scale with your skill level and keep you in control of every decision.