Rules for AI in Your Catalog.

ARCRYN™ permits the use of artificial intelligence in music creation, production, and post-production. AI is a tool, not a shortcut around ownership or authenticity — using it does not exempt you from the same rights, quality, and integrity standards that apply to every release on the platform.

• —AI-assisted and AI-generated music is allowed on ARCRYN™.
• —Use of AI does not guarantee approval — every release is reviewed against this policy.
• —You remain fully responsible for the content, rights, and disclosures of anything you submit, regardless of what tools were used to create it.

• —"AI Generated" — content substantially or entirely produced by an AI system with minimal direct human creative input.
• —"AI Assisted" — content where AI tools contributed to part of the process (mastering, mixing, stem separation, arrangement suggestions) alongside meaningful human authorship.
• —"Voice Clone" — a synthetic reproduction of a specific, identifiable human voice, generated or trained from that person’s vocal data.
• —"Synthetic Vocals" — computer-generated singing or speech not modeled on a specific real person’s voice.
• —"Deepfake" — AI-generated audio or video designed to convincingly impersonate a real, identifiable person without their consent.
• —"Style Transfer" — AI techniques that apply the stylistic characteristics of an artist or genre to new material.
• —"Human Authorship" — the creative decisions, direction, and contributions made by a person, distinct from what the AI system generated independently.

The following uses of AI are permitted without additional restriction, subject to standard disclosure:

The following are not permitted under any circumstances, regardless of disclosure:

• —Unauthorized use of celebrity or public-figure voices.
• —Artist impersonation — content designed to be mistaken for a real, identifiable artist’s work.
• —Fake interviews or fabricated spoken statements attributed to real people.
• —Fake collaborations implying a real artist’s involvement when none occurred.
• —Cloned singers or vocalists used without the documented, verifiable consent of the person being cloned.
• —Outputs substantially reproducing copyrighted training material — including melodies, lyrics, or recognizable samples the uploader does not hold rights to.

At submission, every release must answer whether AI was used:

If yes, you must additionally provide:

• —AI Tool Used — which platform or model generated or assisted the content.
• —Human Contributions — what you personally wrote, performed, arranged, or directed.
• —Vocal Source — whether vocals are human-performed, synthetic, or a disclosed voice clone.
• —AI Percentage — an estimate of how much of the final release was AI-generated versus human-authored.

Every disclosed AI release passes through a structured review before distribution:

• —Manual review — a reviewer checks the submission against this policy.
• —Risk scoring — the release is scored for impersonation, rights, and authenticity risk.
• —Compliance check — disclosures are verified against the uploaded audio and metadata.
• —Distribution — approved releases proceed to DSP delivery as normal.
• —Possible rejection — releases that fail review are rejected with a stated reason and, where applicable, a path to resubmit.

Violations of this policy — including undisclosed AI use — are treated with the same severity as any other content violation. Possible actions include:

• —Warning for a first, minor, non-deceptive violation.
• —Rejection of the specific release.
• —Removal of an already-distributed release.
• —Royalty hold pending investigation.
• —Account suspension.
• —Permanent ban for repeated or deliberately deceptive violations (e.g. undisclosed voice cloning or impersonation).