We’ve Been Warned About Ai and Music for over 50 Years, but No One’s Prepared

This content has been archived. It may no longer be accurate or relevant.

From The Verge:

AI is capable of making music, but does that make AI an artist? As AI begins to reshape how music is made, our legal systems are going to be confronted with some messy questions regarding authorship. Do AI algorithms create their own work, or is it the humans behind them? What happens if AI software trained solely on Beyoncé creates a track that sounds just like her? “I won’t mince words,” says Jonathan Bailey, CTO of iZotope. “This is a total legal c******.”

The word “human” does not appear at all in US copyright law, and there’s not much existing litigation around the word’s absence. This has created a giant gray area and left AI’s place in copyright unclear. It also means the law doesn’t account for AI’s unique abilities, like its potential to work endlessly and mimic the sound of a specific artist. Depending on how legal decisions shake out, AI systems could become a valuable tool to assist creativity, a nuisance ripping off hard-working human musicians, or both.

. . . .

 already face the possibility of AI being used to mimic their style, and current copyright law may allow it. Say an AI system is trained exclusively on Beyoncé’s music. “A Botyoncé, if you will, or BeyoncAI,” says Meredith Rose, policy counsel at Public Knowledge. If that system then makes music that sounds like Beyoncé, is Beyoncé owed anything? Several legal experts believe the answer is “no.”

“There’s nothing legally requiring you to give her any profits from it unless you’re directly sampling,” Rose says. There’s room for debate, she says, over whether this is good for musicians. “I think courts and our general instinct would say, ‘Well, if an algorithm is only fed Beyoncé songs and the output is a piece of music, it’s a robot. It clearly couldn’t have added anything to this, and there’s nothing original there.’”

Law is generally reluctant to protect things “in the style of,” as musicians are influenced by other musicians all the time, says Chris Mammen, partner at Womble Bond Dickinson. “Should the original artist whose style is being used to train an AI be allowed to have any [intellectual property] rights in the resulting recording? The traditional answer may well be ‘no,’” Mammen says, “because the resulting work is not an original work of authorship by that artist.”

For there to be a copyright issue, the AI program would have to create a song that sounds like an already existing song. It could also be an issue if an AI-created work were marketed as sounding like a particular artist without that artist’s consent, in which case, it could violate persona or trademark protections, Rose says.

“It’s not about Beyoncé’s general output. It’s about one work at a time,” says Edward Klaris, managing partner at Klaris Law. The AI-made track couldn’t just sound like Beyoncé, in general, it would have to sound like a specific song she made. “If that occurred,” says Klaris, “I think there’s a pretty good case for copyright infringement.”

Directly training an AI on a particular artist could lead to other legal issues, though. Entertainment lawyer Jeff Becker of Swanson, Martin & Bell, says an AI program’s creator could potentially violate a copyright owner’s exclusive rights to reproduce their work and create derivative works based upon the original material. “If an AI company copies and imports a copyrightable song into its computer system to train it to sound like a particular artist,” says Becker, “I see several potential issues that could exist.”

It’s not even clear whether AI can legally be trained on copyrighted music in the first place. When you purchase a song, Mammen asks, are you also purchasing the right to use its audio as AI training data? Several of the experts The Verge spoke to for this piece say there isn’t a good answer to that question.

. . . .

Even if an AI system did closely mimic an artist’s sound, an artist might have trouble proving the AI was designed to mimic them, says Aimonetti. With copyright, you have to prove the infringing author was reasonably exposed to the work they’re accused of ripping off. If a copyright claim were filed against a musical work made by an AI, how could anyone prove an algorithm was trained on the song or artist it allegedly infringes on? It’s not an easy task to reverse engineer a neural network to see what songs it was fed because it’s “ultimately just a collection of numerical weights and a configuration,” says Bailey.

. . . .

 law will also have to contend with the bigger issue of authorship. That is, can an AI system claim legal authorship of the music it produces, or does that belong to the humans who created the software?

Arguments about whether code can be the author of a musical work in the US are over 50 years old. In 1965, the Copyright Office brought up this concern in its annual report under a section titled “Problems Arising From Computer Technology.” The report says the office had already received one application for a musical composition made by a computer, and it “is certain that both the number of works proximately produced or ‘written’ by computers and the problems of the Copyright Office in this area will increase.”

Despite this early warning flag, current US copyright law is still vague when discussing authorship of works that weren’t created by humans. For now, lawyers are still grappling with the implications of one ruling, in particular, which doesn’t involve computers or AI at all: it’s about a monkey taking a selfie.

. . . .

If a monkey can’t own a copyright, it goes, then what about a song created entirely by AI? Would authorship go to the humans who created the AI, the AI itself, or the public domain?

The heart of this problem is that current US copyright law never differentiates between humans and non-humans. But, the Compendium of US Copyright Office Practices actually spends a lot of time talking about how humanness is a requirement for being considered a legal author. In an internal staff guidebook for the Copyright Office, the Compendium has a section titled, “The Human Authorship Requirement.” There’s also a separate bit to address copyright when a work lacks a human author. According to the Compendium, plants can’t be authors. Neither can supernatural beings or “works produced by a machine or mere mechanical process that operates randomly or automatically without any creative input or intervention from a human author.”

. . . .

Recently, the developers behind Endel, an app that uses AI to generate reactive, personalized “soundscapes,” signed a distribution deal with Warner Music. As part of the contract, Warner needed to know how to credit each track in order to register the copyrights. The company was initially stumped with what to list for “songwriter,” as it used AI to generate all of the audio. Ultimately, founder Oleg Stavitsky told The Verge, the team decided to list all six employees at Endel as the songwriters for all 600 tracks. “I have songwriting credits,” said Stavitsky, “even though I don’t know how to write a song.”

It sounds like a ludicrous outcome, but preventing humans from obtaining copyright on AI-assisted works could limit our ability to use these algorithms for creative purposes. “If you accept AI-generated work as a new form of art and take away the intellectual property rights of the person who created the algorithm,” says Klaris, “you’ve basically said, ‘you’re out,’ and take away their incentive to create.”

Link to the rest at The Verge

PG thinks the potential problems discussed in the OP raise hypothetical questions that most courts will not find terribly challenging.

The monkey selfie case was resolved when the courts held that only humans can own copyrights. Unless Congress changes the law to provide for copyrights to be owned by species other than humans or, per the OP, by machines, a person or persons will own the copyright.

Did anyone claim the manufacturer of the camera owned a copyright to the monkey selfie the camera produced? Of course not. No more than the monkey did.

Additionally, even a very complex machine requires humans in order to create music. A computer program can be protected by copyright with the creator of the program owning the copyright.

When Mrs. PG uses her computer to write a book, she is relying upon physical computer chips that include copyrighted code, a copyright-protected Microsoft operating system, a copyright-protected program to provide a user interface to the system (Windows) and MS Word as a copyright-protected wordprocessing program. While we take this for granted as a simple foundation, there is a very large amount of highly-sophisticated and valuable intellectual property that is necessary for Mrs. PG to create an electronic manuscript that she publishes with Amazon.

Who is the creator of Mrs. PG’s book? Mrs. PG, of course. When the Copyright Office issues Mrs. PG a certificate of copyright for the book, that copyright will give Mrs. PG exclusive rights to the product produced by the process of her fingers pressing keys on a keyboard which begins a process that causes electronic signals to pass through all this proprietary hardware, firmware and software regardless of the necessary contributions of the underlying intellectual property. Of course, Mrs. PG won’t receive any rights to MS Word because she’s not the creator of MS Word.

On a simpler basis, when Hilary Hahn uses her violin, a musical composition by Mozart and a highly-sensitive audio recording system to create a recording of Violin Concerto No. 5, who owns the copyright to the recording? Ms. Hahn and any musicians who played instruments that are part of the recording.

The manufacturer of the AEA R92 ribbon microphone that converted the notes Ms. Hahn’s violin produced into complex electronic patterns in the form of digital representations of lengthy strings of numbers that then pass through amplifiers and a variety of other enormously sophisticated electronic devices which modify the lengthy strings of numbers in a wide variety of ways before those numbers are used to create proprietary patterns of electronic charges that are fixed as patterns of charged particles on a magnetic medium that is has been affixed to a disk or incorporated into a variety of electronic chips. (Yes, PG knows that sentence would be difficult to diagram.)

All of these hardware and software devices use proprietary intellectual property that belongs to people other than Ms. Hahn to accomplish their purposes. Ms. Hahn gains no copyright to the creations of the humans that created the hardware and software.

When someone wishes to listen to one of Ms. Hahn’s performances, a host of proprietary hardware and software that is both far from and near the listener accesses an electronic representation of Ms. Hahn’s performance, converts those electronic patterns into other numerical sequences that can be transmitted via the internet. When those numbers arrive at a location near to the listener, another set of hardware and software transform those complex electronic patterns back into vibrations that can be discerned by the human ear and the listener enjoys an audio experience akin to that which the audience at Ms. Hahn’s live performance enjoyed.

Ms. Hahn owns the copyright to her performance all the way through this lengthy chain. Although the violin actually produced the sound waves everyone loves to hear, the violin does not own the copyright.

PG submits that for copyright purposes, music created by a computer that includes AI software gives the AI no more ability to own a copyright than a stereo speaker has to the vibrations it produces to create the sound of Ms. Hahn’s performances in PG’s living room. Anything the AI system produces is the creations of the humans that caused the AI to create the music. The AI system is akin to Ms. Hahn’s violin, a non-human device that requires a creator’s input to produce unique musical sounds a human can hear.

18 thoughts on “We’ve Been Warned About Ai and Music for over 50 Years, but No One’s Prepared”

  1. Quibble on one aspect: Ms Hahn and the musicians do not own the copyright in the recording… unless the actual person who did the recording did it as a work for hire for Ms Hahn. This is one of the counterintuitive problems with copyright law: Copyright in a fixaation belongs to the person who fixed (recorded, painted, handwrote, typed, etc.), not the performer, unless there’s a specific agreement otherwise that precedes the recording. (As an aside, this is one of the many complications in the King family’s assertions of rights over the “I Have a Dream” speech, but we’re getting into law-school-exam-question hypothetical territory here…)

    So the preliminary question for AI-created music is whether the AI’s output constitutes a fixation of the music, or merely directions to create a fixation of music. And I’m not defending copyright’s “fixation is what matters” meme; it’s just what we’re stuck with in the law as it exists, and not just in the US.

  2. The work around people will use is simple–the A.I. will create the work, the person who owns the A.I. will register the copyright.

    One problem is that there have been lawsuits with music being judged too similar to another work. Whoever, controls/programs the A.I. is going to want to have a method of adding the ability to change small parts of the music so it no longer resembles the song(s) it mimics so that they can claim its a totally new work.

  3. Beyonce also has trademark protection in that she has a distinctive voice, and trying to copy it leaves them open to an infringement suit. This has been successfully done when commercials hired “sound-alike” personalities to mimic the celebrities they couldn’t get.

  4. I would be interested in RobMozart or other classical projects. Instrumentals are a natural. I wouldn’t be surprised to see Karl Jenkins jump on this. It’s at core an evolution of the music synthesizer.

    No need to stomp on the living when you can play in other fields.

  5. Since true AI doesn’t actually exist yet, and barring some sudden leap in tech won’t for some time, it’s a hypothetical question. Software is owned by the creator. Programs designed to make music, regardless of how advanced they are, are only doing what they were created to do.

    • The original sentence reminds me of some German news stories I’ve read. 1. Find subject. 2. Find verb. 3. Read 10-15 modifying clauses between subject and verb. 4. Read other two clauses following the verb.

      Hmmm. Maybe all lawyers are descended from a long-lost German-speaking common ancestor? 😉

      • It’s a possibility; certainly many, many lawyers have an inordinate fondness for dependent clauses — especially subjunctives — when drafting material, including contracts of adhesion for nonlawyers, in part in the not-forlorn hope that careless readers will lose patience and fail to discover the verb at the end of the sentence where it belongs.

  6. The is no problem with AI and music. The problem lies with the forever and a day copyright laws. Not that rights holders want to hear that their greed is the actual problem.

    I’ll link once again to Spider Robinson’s ‘Melancholy Elephants’

    https://www.baen.com/Chapters/0671319744/0671319744.htm – first chapter.

    It boils down to the minor fact that there are only so many ways to make a tune that doesn’t sound similar in some way to another tune.

    The same of course goes for stories, but we have more than 88 notes to string together in different ways …

    • Agreed, A.

      The lengthy period of time during which a copyright persists is of primary benefit to large corporations, Disney, for example, who have acquired copyrights from the original creators.

      Walt Disney is an exception to the general rule of the creator receiving only a small portion of the financial benefits of a copyright because, in addition to his creative genius, he was a management and promotion genius as well.

      For authors who grant rights to their creations to large organizations for the full term of their copyright, the large organizations receive the lions share of the benefit of the fruits of the copyright and may even be in a position to prevent the author from creating derivative works based on their original creation.

      • Even more fun is those DCMA take-down AIs. NBC or one of the other big ones post something and then has their bots throw take-downs at anything like it. I remember a bit of a stink when NASA couldn’t post their own pictures because of someone else posting them first. That’s one of the laws that’s in bad needs of ‘fixing’, along with that ‘death+70’ or whatever it’s up to now.

    • Thank you so much for that story link — I have been looking for this story for decades, but could recall neither the title nor the author.
      Being married to an Intellectual Property lawyer makes it especially important to me.
      And now I can read the rest of the anthology as well!

      • I’ve trotted it out a time or three now. It’s funny how often the latest ‘big bad thing’ is far from new and has been warned against – only to be ignored because it wasn’t the answer/message the powers that be wanted to hear.

        MYMV and you find what you seek. 😉

  7. I see no problem with A.I. and copyright for now. But down the road when our robot masters assume control of the government…

  8. says an AI program’s creator could potentially violate a copyright owner’s exclusive rights to reproduce their work and create derivative works based upon the original material.

    Few people create AI programs. Many more use them. An AI program based on neural nets can be purchased untrained. Then I can train it on Mozart, while someone else trains the same program on Michael Jackson.

    Then I can add Elvis, Kate Smith, and Mario Lanza to my Mozart. It goes on and on…

Comments are closed.