Most people still do not understand why AI music puts lawyers in a bind, so here is the real issue.
On suno.com, you can sing every part of a track. The melody. The harmonies. The beat box rhythm. The bass line. The orchestral parts. Suno digitally analyzes your voice and turns each sound into a full production.
Example:
https://www.youtube.com/shorts/Dt2nc_S0hz8
You created the musical work.
You performed every element with your own voice.
You dictated the arrangement and structure.
Yet the US Copyright Office will only let you register the underlying musical composition. It will not let you register the sound recording that Suno outputs, even when the entire creative content came from you.
Here is the legal backdrop.
Thaler v Perlmutter, 2023 WL 5333236 (D.D.C. 2023)
The court held that a work with no human authorship cannot be copyrighted. The Copyright Office now uses this to deny human authorship in AI rendered audio.
Naruto v Slater, 888 F.3d 418 (9th Cir. 2018)
The Monkey Selfie case. Only humans can be authors. The Office treats AI systems like the monkey. If the machine renders the final output, the human loses the recording.
And here is the ghost that still haunts all of this.
White Smith Music Publishing Co. v Apollo Co., 209 U.S. 1 (1908)
The Supreme Court held that a piano roll was not a copyrightable copy because humans could not read it directly. Congress later fixed this with the 1909 Act, but the mindset survived. If a human cannot directly perceive the final form, the law becomes suspicious of authorship.
That thinking is still baked into the Copyright Office today.
If the machine is the one that renders the final audio, the Office treats the human input as somehow incomplete, even when the human supplied every creative element.
Which leads to the question no one is asking:
What happened to derivative works
Under 17 U.S.C. 101, a derivative work is a work based on preexisting material that includes new human authorship. That is exactly what is happening here. You supply the melody, harmonies, rhythm, arrangement, and creative direction. The AI is simply transforming your human authored material into a different medium.
By any traditional analysis, the resulting audio should be a derivative sound recording with human authorship.
Yet the Copyright Office says the answer is simple:
Not you.
This is the authorship dilemma that is about to explode.
Reactions anyone?