Skip to main content

SC-054 · Expert

AI and Music Industry Legal Challenges

Guest: Jeff Liebenson, Entertainment Lawyer and President of the International Association of Entertainment Lawyers

Summary

Jeff Liebenson is an entertainment lawyer based in New York City and President of the International Association of Entertainment Lawyers. He started in the physical music era and now advises startups and companies on IP rights and licensing. He says AI could replace artists in some background uses, but it could also help songwriters generate drafts and collaborate.

Streaming raised total consumption but cut per stream royalties and broke the album into singles. Liebenson explains that AI is now the next disruption, with legal fights over training on copyrighted music still unresolved in the US and Europe. Rights holders must choose between waiting for court rulings that could maximize damages or striking voluntary deals.

He also compares US and European licensing: American companies must clear mechanical and performance rights separately, while European societies often issue joint licenses. His practical advice for founders is to design products with licensing in mind and balance protection with the opportunity AI offers.

As of the episode's release on 30 May 2024.

Key takeaways

  1. 01Jeff Liebenson sees two big AI risks: mass replacement of working artists and legal uncertainty over training on copyrighted music.
  2. 02Streaming lowered the barrier to entry but spread subscription revenue across so many songs that royalties are often pennies.
  3. 03Rights holders who wait for favorable court rulings could win maximum damages, but voluntary AI deals provide more certainty.
  4. 04The US split mechanical and performance rights across different societies, while European societies often issue a joint license.
  5. 05The Andy Warhol fair use ruling focused on market effect, which could weaken AI companies' fair use arguments.

Chapters

  1. AI concerns and episode intro
  2. Jeff Liebenson's background
  3. The shift from physical to digital
  4. How streaming changed royalties
  5. AI threats and creative opportunities
  6. AI investment and legal risk
  7. Regulation, deals and negotiation strategy
  8. Fair use and the Warhol case
  9. Licensing in the US and Europe

Guest

Questions this episode answers

How could AI replace or help artists and songwriters?

Jeff Liebenson says AI generated music is already good enough for background uses, so many working artists fear being replaced. But AI can also act as a co-writer, generating first drafts and ideas that a human songwriter then accepts, rejects or builds on. He sees both a real threat and a major opportunity.

What should music rights holders do about AI training on copyrighted music?

Rights holders who want maximum payout can wait for favorable litigation rulings and then seek damages plus ongoing licenses. But Liebenson notes there is no guarantee of winning in court, and many parties prefer to negotiate voluntary deals to remove risk and secure a satisfactory result.

Why is the Andy Warhol ruling important for AI music cases?

The Supreme Court found Warhol's use was not fair use because it competed in the same market as the photographer's original. Liebenson says AI companies training on music and releasing AI tracks into the digital music market could dramatically affect that market, which weakens their fair use position.

Why is music licensing more complicated in the US than in Europe?

In the US, mechanical and performance rights are handled by separate societies, so digital music requires dual licensing. European mechanical and performance societies created joint ventures that issue one combined license, which Liebenson considers a saner approach for companies that need rights.

AI can generate, you know, music that's not bad right now. It's going to improve. It's going to, I'm sure, generate lots of wonderful music in the future.
Jeff Liebenson

Episode notes

Artificial intelligence is reshaping the music industry, and with it comes a wave of legal and creative challenges. In this episode, entertainment lawyer Jeff Liebenson delves into how AI is influencing the music scene, offering both opportunities and risks for artists and creators. 

Understand Jeff's insights on managing the complexities of copyright, balancing innovation with protection, and the broader implications of AI for the future of music. This conversation provides essential knowledge for anyone navigating the intersection of music and technology.

Highlights:

  •  The shift from physical to digital media and its economic effects.
  •  AI's role in both replacing and boosting artists.
  •  Legal complexities in AI-created music and copyright conflicts.
  •  Approaches for using AI while protecting creative rights.



Topics

Transcript

Transcribed automatically. Names and terms may be misspelled. Every line is timestamped: select a time to play from there.

Read the full transcript

Well, I have two main concerns. One is that it could destroy everything for everybody. Other people, deeper thinkers than myself, think it could be the end of the world. It could be the end of humanity. The other problem I see is that AI could be a fantastic boon to artists and the industry. And you could use AI to generate thoughts and first drafts. And you can sort of team up with AI. And, you know, instead of working with a co-writer, a single songwriter could theoretically work with AI. Music Technology Conference in Bergen is close. And we're doing a partnership episode with them. We have Jeff Lievenson. He is president of International Association of Entertainment Lawyers. We talk about the legal aspects of what used to be in the industry and what is to come. It's a very interesting talk. Hi, guys, and welcome back to the Sound Connections Podcast. Today, we have a partner episode with MCC Bergen, which is a new technology conference that happens in Bergen, Norway that I was a part of last year. And Jeff, you were also

there. Welcome, Jeff. Hi, Jacob. How are you doing? I'm doing well. I met you last year, and I'm very happy to see you again this year at the conference. The conference is quite interesting. It's sort of last year as a gathering of a bunch of great people that sort of just came together and shared some thoughts on sort of current topics. And it's actually one of the best networking spaces I've been in for some time because of the quality of people that are there. And Jeff, today we're going to talk about you and your career and your domain expertise with being an entertainment lawyer. But for people who don't know you, could you maybe explain who you are and what you do? Sure. My name is Jeff Lievenson.

I'm an entertainment lawyer based in New York City. I've been in the music field in particular for a long time. Worked in the industry itself. Worked at a major record company, BMG, before it merged with Sony. I've been in a couple of large law firms in New York. And now I have had my own practice in New York City Lievenson law for almost 15 years. And it's been a lot of changes in the industry and a lot of changes in my personal involvement of things over the years.

I can imagine. This is an industry that's been through a lot of change and consistently over many, many decades. Like it's every single decade there's a huge change that the industry needs to adapt to. And you've seen your fair share of those. There's a lot of things happening in the industry right now that I'm very sure that you are very busy with. But before we get into that, could you just walk me through what are the major changes we as a music industry on an entertainment lawyer's side that has happened the last couple of decades that's made like the biggest shift that you needed to work with?

Everything has shifted. It really has. When I started, it was all about physical product in the music industry. It was about the artists and recording the artists and shipping physical products, CDs and cassettes even, and vinyl. And so the relationship between the industry and the customers is really like a B2B business. We dealt with the press. We dealt with retail record stores. And we dealt a lot with the artists.

And when digital happened, it really endangered the industry. It didn't react by embracing it. And gradually over the years, as technology has taken its hold, the industry has gradually and increasingly embraced it more and more. But it's a real major shift from a physical format industry to a digital format industry. There's still a little bit of physical, but the bulk of the income and the bulk of the industry's interest is definitely in digital.

And while that has happened, there have been so many seismic changes because the album, which was the key configuration, has really been broken up. There's still albums, but almost all consumption is track by track singles. That totally changes all of the economics. It changes the approach of the A&R executives that work with the artists. It just changes everything. And I think in the old days, we would have just loved the idea that everyone would have a stereo system with them everywhere they go.

We had to wait for people to go into record stores and then bring the record or CD home and put it on their sound equipment. And now everyone has the sound equipment in their pocket. So everybody's a potential consumer. And I think the market is so much faster. It's so much more worldwide than it ever was for so many different types of music. So there are many, many wonderful things that have happened with the digital transition. Some are really challenging for the industry. Some are really a godsend. So it's navigating all of that.

What would you say would be the single biggest pain point of, from a legal perspective, the digitization change that happened? Rates have gone down for artists and songwriters. I mean, the rate, I shouldn't say that. As a careful lawyer, I shouldn't say that. The rate is insane, but the dollars and cents at the end, the amount of the royalty payment they receive is much less. You know, we're not selling albums for 12 or 15 or $20 or whatever your local currency is. And that's where the album was bundled together and people were buying all of the songs and what now look at like relatively high prices. Now people are paying their monthly fee for their streaming service of choice and they're able to have unlimited consumption. I mean, it's the greatest time ever to be a consumer of music. You can listen to anything you want. It's at the press of a finger. You can be up on any music you want. You can dabble. You can experiment. You have access to

everything. And you pay probably less than the cost of a CD every month. So the pool of money, in one way, it's much less, I think, because active music consumers have to pay less. If you've ever bought more than one CD a month, you're probably paying less. But the market correspondingly is larger because more and more people are consuming music than ever before. So that helps to balance it out. But as everyone knows, after all the money goes in the pod and then it gets divided based on the number of streams or whatever the methodology is, the amount that comes out to the artist is, you know, it's not what it used to be. And especially if you're a songwriter.

Yeah. Does that come from the back in the day, if you were a songwriter that were releasing albums, you had reached a point in your career that was sort of unattainable for the general public. But now when sort of the barrier of entry to being creative of having some sort of response from the industry on, let's say, the pro rata model where you could gain significant amount of streams, maybe not having the same level of craftsmanship, that you need to share that pie. Is that is that sort of the lower barrier of entry for creators that's taken part of the economy? Is that the biggest factor now would meant sort of the generally speaking that the revenue of the music industry has reached? It's what it used to be.

That's part of it. I think it's several things. You know, the monthly fee that people pay for a major streaming company. It's only so much money. And when you divide it up among all of the all the record labels on behalf of the recording artists on behalf of all the publishers, I'm sorry, to all the publishers on behalf of the songwriters, and then it gets whacked up. There's so many millions and millions and millions of streams that once you divide it, you know, you're talking about pennies. So, you know, a songwriter could have a song streamed a million times and still have a very small payment. So that's part of it. What you mentioned is part that the barrier to entry is down. If you want to write a song and record it, you can put it up on the streaming service. So you're there and maybe you're brilliant, maybe you're not, and you're in the service. So you may not generate a lot because there is so much music out there and it's so hard to rise above the fray that, you know, for that reason, it's hard for you to even get a big share of the pennies

that are out there. But it's also been a, you know, a long battle that the music publishers have been in to get, you know, their fair share of all the revenue. And, you know, they're battling and they have raised the rates for music publishers and songwriters quite considerably. But they would be the first to tell you they still don't get an equal amount to the record labels. And if they did, it would be more than a hundred percent of the revenue that the streaming service gets. So it makes it, you can't share more pieces of the pie than there is pie. And so unless people are going to give up part of what they receive, you know, the question is, where's it going to come from? So it's a challenging issue.

I'm going to do something dangerous, Jeff, and that is I'm going to present the working theory to a man that knows a lot more than me and I'm ready to be butchered. But when you talk about sort of the negotiation between the publishing share and the master share, obviously a lot of the major companies, major labels have also publishing sides. But is there generally speaking a lower interest in having publishing share because the journey, the cut to the creator, the publishing side is higher than the master. Just to sort of reiterate like a, you know, a normal, let's say Scandinavian cut would be 33 to 50% on a publishing deal, while a major label deal would be closer to, you know, 80%, 90%, if that makes sense.

So the incentives of having a bigger share of the publishing would result in a lower revenue to the group at large. Is that a thing that actually comes into play or is it just me speculating? Yeah, I'm not sure I see it quite that way. But, you know, some people think that the, since the major publishers are owned by entities that also own record labels. Some people speculate that the corporations are more interested in seeing more money on the record side.

And, you know, I'm not sure that really is true. I mean, you know, I understand the perspective and the reasons for it. But the people that work at the publishing companies very much want to earn their bonus businesses and increase their revenue. And, you know, I just don't think you see instances of the record company picking up the phone and telling the publisher to cool their jets and not take so much because they want it on the record side. And I think both the record company has a fiduciary obligation to their entity and the artists they represent. And the publishers have a fiduciary obligation to do as well as they can on their side. So they're usually really competing against each other, and arguing for as much as they can get on their own side.

The rate that's set right now, and I know we're sort of diverting a bit from the main core topic, but this is law. So it's really interesting. This is the negotiations between parts where you were sort of involved. But the initial rate that was set in between, was that just like where they started and they were sort of stuck in that setting? There's something to that. Because when digital started, everyone thought, okay, how much should we pay?

Who should we pay it to? And I was involved in some of the earliest discussions. And people came to me and said, oh, you know, how much should we charge for a download of a single? And, you know, I thought it was interesting that they were asking, but, you know, where do you start? You know, we're at the beginning. This was at a stage when no one was sure that anyone would really pay for digital music. Until iTunes came around, I worked on the first iTunes deal. You know, no one really knew for sure if people would pay for digital music, for a digital file. It was something people weren't used to.

And so how much would you charge? And I think people, this is how I looked at it, and a lot of people, is you look at historical models. Well, how much do you sell for a physical CD? And who gets paid the royalties on a physical CD on the recording side and on the publishing side? And then you say, well, it's not a physical CD single, but that seems to be one of the analogies. So you look at that, and, you know, that's like a starting point or, you know, just a reference point.

And then you say, okay, how is it similar to that, and what are the ways it's different? You know, you're not manufacturing, you don't have trucks shipping them. You know, there are a lot of differences. And, you know, I think you sort of start there. So part of it is history. You know, that this is the way things have been done before, and that, you know, it's rare in any industry that everything dramatically changes as if there was no past. That people start with a totally blank slate.

So when digital started, the record industry started with all of its experience and that influenced where we started. It's sort of been a while since digitalization really sort of disrupted the industry, but now we're going into, you know, a new technology advancement that has something that seems to some degree similar in disruption, and that is AI. What is your primary concern with the introduction of AI as new technology in the music industry?

Well, I have two main concerns. One is that it could destroy everything for everybody. I mean, other people, deeper thinkers than myself think it could be the end of the world, it could be the end of humanity. I like to think smaller and just think it's at the end of our music industry. And I've talked to a lot of artists and a lot of songwriters that are very afraid that AI could replace them. AI can generate, you know, music that's not bad right now. It's going to improve. It's going to, I'm sure, generate lots of wonderful music in the future. There's definitely a big difference between an individual human being artist who has deep feelings of things. I mean, artists are brilliant people. They have a way to observe the world that most normal people don't. And not only can they observe it, they have the ability to communicate it and articulate it in a beautiful way. And that's a

rare gift. That's why artists are so special. But, you know, are they going to be competing with computers and training models and generative AI and all the rest? I'll tell you one thing, generative AI is faster. Generative AI is more prolific. The quantity it can disgorge. And for many purposes, does it really matter if it's a heartfelt human emotion in a song or not if the music is used for, you know, background music in a mall or, you know, I mean, there's an awful lot of uses of music today that are background.

You know, people don't, you know, when I started, you sat between your two big speakers and you spent the entire time for an album and you got into every detail. And I think, you know, that's something that some people still do, but many people, you know, they've got their phone in their pocket and they're walking down the street and they're looking at the traffic and doing other things and multitasking and music isn't always the front and center thing that it used to be more than today. And so we still have that experience for sure, but we also have a lot of background experiences. And is AI music going to be viewed by some people as sufficient for that? You know, the main music industry has a lot of what they call production music houses now where they have vast troves of background music and, you know, mood music and things that are good just to set a mood and they're not deep heartfelt human emotional messages. So, you know, I think that's probably the most vulnerable part, at least initially, but,

you know, what's going to happen in the future? So that's a big problem. The other problem I see is that AI could be a fantastic boon to artists and the industry. I'm not personally an artist, but if I, you know, I'm sure artists sort of draft out their ideas, like they try a song or they have a, you know, a pencil and eraser and they try different thoughts and they're erasing and they improve and iterate. And you could use AI to generate thoughts and first drafts and, you know, you can sort of team up with AI and, you know, instead of working with a co-writer, a single songwriter could theoretically work with AI and bounce ideas around and, and the human artist could get input that he could then, you know, in his own judgment, decide to accept or reject or say, oh, then, you know, that gives me an idea. And it could be off and running based on some thought he got from the AI. I know that's how it works

often when you have two co-writers in a room. They, they inspire each other and, um, right back. Yeah, I guess it's probably better to have two co-writers, but if you don't have a co-writer around and you have AI, you might work with them, work with it. So I think there's a big opportunity and also there's opportunities for the big corporations and music to streamline, uh, certain processes and get better things and get more information and more data quickly with AI. So it can improve. You know, I think there's a lot of opportunity to improve the industry. And one concern I have is that we get too concerned and too fearful and try to shut down more than we need to and not be able to avail ourselves of the upside. It's a delicate balance. We don't want to take over, but I'd like, I'd like to be able to use it wherever it's good for us. Yeah. Because you, you might argue that one of the mistakes the industry at large did, uh, last time around was being too reluctant and too resistant

and left out a good deal of time that could have been provided more revenue. Um, is that sort of how you see that this is a more, let's call it a wise response to technological innovation and rather seek the opportunity than try to, uh, resist? I think if a lot of this is put in the laps of the lawyers in the business, you know, the top executives often, you know, really want the lawyers to get deeply involved in us and they ask us what we think they should do. And lawyers are all different. Some are very, very conservative and protective. And one of the key things the lawyer does is protect his clients.

So that's totally appropriate, but a lawyer also wants to help his clients seize opportunities and, you know, achieve advantages. So, you know, it's a balance how you do it. And so some lawyers tend to see it a little more protectively and others tend to see it a little bit more. How can we get the best out of it? Um, and, you know, personally, I've always been one of those saying, how can we turn it to our use? How can we get the best out of it? While almost trying to cover the downside and have protection where you need it. It's, it's a delicate balance.

And, um, you know, when digital started, there was a lot of risk to the very lucrative physical market. You know, it was risky saying, let's embrace it. And at every stage in the digital transformation, there've been things to legitimately be concerned about and legitimately to try to take advantage of. And it's a balance, you know, how, how do you, everyone has to strike their own balance. And I guess the industry at large strikes a balance, um, between protection and embracing.

Hmm. Um, it's only fine if you don't comment directly on the case, I'm going to present you, uh, you can also answer it in more generic nature. But if you look at the recent, uh, round, uh, series B, I guess, from, from Suno, uh, that did $125 million. One, a lot of people would argue that there's sort of a whole new industry and worth that towards consumers that are being most likely, uh, based on rights that other people have created previously. Uh, in other words, there might be, um, an opportunity for extra revenue towards music that has been trained on existing rights that maybe is not being, um, compensated for. How would you sort of value that, um, that conflict? Like what, what would be your perceptions on, on that conflicting movement that's happening there?

Well, I think the overall perspective is there's a lot of money flooding into all types of music right now. If you're a, you know, a great songwriter who's, you know, very famous and has a huge catalog, there's an awful lot of money coming into the industry to acquire those rights. Money's coming in from the finance world and it's coming in for a lot of reasons, but people are getting, you know, payments they never would have gotten say 10 years ago. So there's money coming into music in general. Uh, with Suno, we're talking about money coming into the AI industry.

And right now, you know, people are speculating is AI a bubble, uh, for investors or not. But the plain fact is there's enormous amount of money going into AI. It's the, the top flavor of the month, the flavor of the last two years. Everyone wants to get in on it. The company's values are just growing going astronomically. So I understand people being very, very excited about it as potential and wanting to get in now while they still can. And they see great riches ahead, um, with a company that's training on copyrighted content and then generating music from that. There's a big risk. There's a big unknown thing. Because it's possible that will be viewed as legal and it's possible. I may be viewed as illegal. And if you invest all that money in a company that's engaged in illegal activity, I don't think things will work out very well, but it's very uncertain right now how that'll go. Uh, there are,

um, you know, in Europe, there's the new EU AI law, which said that copyright must be respected. But it also says that copyrights have to be protected unless there's a, an exception or limitation under the copyright law. And one of the exceptions that copyright law in Europe is, uh, sorry, there's some background news, uh, background noise here. Um, is if there's temporary copying going on, that sometimes is viewed as an exception to infringement. So would that apply to AI or not? I can easily give you a good argument why it is or why it isn't. Uh, and that's probably, you know, that could come down to what the courts ultimately decide. And as you know, courts move very slowly. So we just don't know how that's gonna play out.

And in the U S we don't have all like the EU AI law, at least not yet, but we have several litigations that are going on both in the music industry and in, uh, journalism with the New York Times and with book companies and, and all sorts of different types of media. And each of these cases are a little different and the differences could really affect the outcomes. But those cases, you know, AI's really been front and center in the public mentality since around November of 2022. So it's not very long. And these court cases started several months ago, but they have a long way to go before So we don't know what the law will be proven to be whether training on a, on copyright content is illegal infringement, which is the contention of all the content owners, or whether it's fair use in America or fair dealing, uh, as it's called in some other jurisdictions. Um, you know, and that's going to

really dramatically affect these companies like, uh, Chad GBT, which is apparently just trained on a lot of copyrights. It's a very interesting future going ahead. Um, and I, I want to understand regulation obviously, uh, is slow. Um, but it's also very impactful. Um, how do you see this moving forward timeline wise? And I, I know this is a big ask, but major regulations that will protect the rights of the original creators of music. When would that be put into place effectively? So there's a good structured regulatory environment for this to happen in Well, there's two ways that could happen. We can wait for the court cases to play out and we can wait for the European, you know, the EU's AI law to be implemented in each of the

various countries, which is going to take a while. You know, then we'll have a clearer sense, but that's going to be years. The other way it could happen is through voluntary agreements. It's possible that if the AI companies want to come to the table with the rights holders and cut some deals, that they will voluntarily reach agreements on what the rules of the road are, how much they should pay to train on copyrighted content, what they should pay for it, how they can provide record keeping so that the rights holders know how to allocate the money to their, uh, creators. Um, and we already see that happening to some extent. Um, you know, some of the companies are cutting deals right now. Um, but it's, you know, they're really exceptions. Um, I think a lot of people are really waiting to see how the court cases play out. So they know if they have a stronger hand in the negotiation or a weaker hand.

Okay. Uh, it's, it sounds a bit like gambling, uh, to understand sort of where, where that will end out. Uh, what, what is your, um, sort of initial understanding of where the, that's called the, the maximal worth for the music industry. Is that through most likely pursuing deals or waiting for a regulation? Well, you know, you just said it's a little like gambling. Um, if you want, and my advice would be if people want to go for the maximum payout, if they want to get the most they can out of this, then they ought to wait and hope for favorable across the board litigation rulings to put them in the driver's seat. And then there'll be the, you know, they'll, they'll be able to get all the damages from all the massive copyright infringement plus enter into future ongoing license agreements. That would be

the home run economically. Uh, but there's no guarantee that's going to happen. It may not play out that way. You know, some of the early rulings aren't that favorable. Um, you know, but it's very early days. It's hard to read the tea leaves right now. Um, and, uh, what usual, what happens in a lot of fields outside of music is, you know, there's a case in court and people are evaluating how comfortable do we really feel we're going to win? Because at the end of the day, even if you have the strongest case possible, there's no a hundred percent guarantee you went in court.

You know, there could be something strange happening or the judge gets out of bed on the wrong side of the bed or, you know, there's unintended things sometimes, unexpected, unpredictable things. But, but if you feel highly confident, then you dig your heels in a little more. If you think it's 50 50, you play accordingly. If you think we have, uh, you know, a less than 50% chance, then you negotiate, but not from such strength. Um, and I, you know, oftentimes in many, many fields, and even in the music field, people feel it's better to enter into a deal and get a satisfactory result, even if it's not a hundred percent of everything because it gets the risk. You said that there's been some early court rulings that maybe have not been in favor of, uh, the music industry of the creative industries. Who would you say right now has the upper hand in the current, um, fight or whatever you will call it?

I think in the AI cases, uh, I don't think anyone has the upper hand. I think that, um, both sides are really well funded. I think both sides, you know, have a certain conviction about the, uh, you know, about their positions. I can't imagine the rights holders giving up. I mean, this is vital to them. Um, and, uh, I think it's, it's more vital in a way to the rights holders because, uh, because of the consequences and the AI companies, we just talked earlier about how much investment is going into them and they're training on an awful lot of content, copyrighted and non-copyrighted.

And there's a lot they can do outside of the copyright field. So I think this is like, you know, one part of, from their side, it's one part of their business. And if they pay for it, they've got plenty of money. So I don't think it says all or nothing kind of feeling on their side. I think they'd like to win and they'd hate to have to deal with all the rights holders. But if they had to, um, yeah, I think I would like to think they'd be willing to come to the table when the time is right. Hmm. I was going to ask another difficult question. Um, and as a lawyer, you've been through a lot of cases and you've seen a lot of these aspects, but, but as a, listen, as an ordinary person that, um, sort of come from music, um, and I look at these technology investments, let's just take Suno.

You know, we, we would just assume that they're training on, on copyright. They, they do not own. Um, the founders and the team must have an approach of making conscious decision of violating some rights. It, what is the genuine, not just with Suno, because we don't know what that case is, but in general, your experience with these type of disruptive companies do is the sentiment that it'd do it on, on sake of technology advance enhancement, that's going to help humanity.

And that's sort of how they sort of wiggle, uh, they, they try to navigate that complex emotional landscape of you might be hurting someone's living from this. Like, what is the, what would you say would be the typical arguments in such a, uh, ethical, um, internal conflict that might happen in some of these technology companies? Well, if they sat down and had an ethical, ethically focused conversation, um, what they might be thinking is that they're generating a lot of, you know, AI generated music. It can be very valuable. Um, if it competes against human created music, uh, you know, that's sort of what happens when technology advances, when automobiles came in, it wasn't so good for the buggy companies that were in existence. Uh, you know, when, you know, when personal computers came in, wasn't great for the portable typewriter companies. So, you know, new technology often just

steamroll us over preexisting, um, practices, business models, technologies, they get left by the roadside. And, you know, so if you come up with the new technology, um, you know, do you, you know, cry a little at night for the people left by the roadside, or do you just revel in your advances and how you're helping society move forward? And, you know, hopefully it's a balance and an ethical work. Um, and you know, I think the, the issue, if you're, if you're, if you're an AI music company, you can focus on the generated works, which if they do it right, really shouldn't be substantially similar to any, um, existing copyrighted music, but the training is where there's significant copying of copyrighted music. I mean, it's possible to have a generative system that doesn't train on copyrighted music, but you know, all the ones that we know that do

train on copyrighted music, you know, they're, they're feeling, and I think probably what they think ethically is all we're doing is looking at the relationships within the music. We're not, you know, lifting some phrases or cop, you know, engaging in copying. We're just, you know, looking at the relationships within the music, feeding that into our AI machine. And then our AI machine learns from those relationships. They think that's facts that they're copying, not copyrighted content. And you're allowed to copy facts under the U S copyright law.

Um, you know, I think what they're really doing is by copying the whole work and it's not just facts because they have to ingest, you know, full copy of the work, but that's what the courts will have to look at. We had a couple of big cases in America where they used the entire work, but they used it for a transformative purpose. There was a big case involving, uh, Google when they had their, their Google. I remember the book one. Yeah. They copied the whole book and the court said, well, that's all right. And everyone was a little surprised at first, but they, they said they're making a transformative use out of it. And a perfect 10 pictures where they copied the entire photos from these magazines and they use them as digital pointers, like, you know, like search links. And they said, well, they're not using the picture really as a picture. They're using it as a search link. That's transformative. It's a little more involved than that, but they say, no, they copy the entire work.

It's put to a transformative purpose. You know, I, I think that we just had another big case involving Andy Warhol here, a copyright case, the one to the U S Supreme court. And there are four factors under our copyright law that are listed in the statute. And one is whether the, you know, the nature of the use and that gets into whether it's transformative. But another one of the four factors is the effect on the market and the U S Supreme court really focused in the Warhol case on the effect of in the market because the Warhol work, which was taking a photographer's photograph of prints, the artist and putting color shapes on it and making it an Andy Warhol piece. Um, he was using it in that case, um, to license it to a magazine for use on its cover. And the photographer whose picture was taken, his main business was licensing his photographs for use by magazines on their covers.

And so they said it's the same market and it has a big effect on the market. And they looked at all four factors, but they said, that's the factor that seems to tip the balance here. And so they found that the Warhol was not a fair use. And you know, this is the same Andy Warhol who has, you know, paintings up in magazines and museums around the world of the Campbell's soup can, which is almost that copy of the artwork on the can. And everyone thinks that's fair use. And he's never, there was no liability for that. Um, but here with Andy doing what Andy does, there was, it was found not to be fair use.

So I think that's not a great case for all the AI companies because what they're doing training on copyrighted music and then pumping all this AI music out into the market and trying to exploit it could dramatically affect the market for digital music. So we'll see how the courts evaluate it all. It's very interesting. And, and sort of, I don't want to give too much in this episode to the audience because they need to come to the MCT burger and hear you speak. Uh, you're going to bring a lot of value to the table there. Um, what, one interesting thing sort of to, to add to the complexity of it, uh, is sort of also the difference between American law and European law per se, like the way that they fund it fundamentally work differently. Um, and it's really interesting to see the movements of both markets and then sort of see the consequences of that. Yes, definitely. And that really gets into what I do every day in my practice. We've been talking a lot about litigation and AI. It's a lot

of it will be, you know, it's being addressed right now in these pending litigation. So it's very important to be up on that. And some of my clients are, you know, very focused on that. But most what I do is deal making and licensing. So I'll advise, you know, startups and other companies, what IP rights do they need? How can they help to design their products so that they have a product that they can license or maybe don't have to license? Uh, some things in digital music are easier to license than others. And so, you know, having that in mind as they're proceeding, because often business models and technology changes. And so you can change in a way that makes it really, really difficult to get the deals you need or, or make it easier. So I have a lot of conversations with clients about that, and then, you know, put them in touch with the right people and have the right conversations to try to get them the deals they need. That's a fundamental thing that I'm involved in. And,

and your question really relates to that because in the US, the people you have to go to for the rights are different than they are in Europe. And, you know, for, and this goes back to what we said earlier about our things affected by historical practices. A lot of this comes from historical practices. Um, and you know, there used to be an approach before digital, where if you had, if you were putting out physical products, CDs, for instance, you had to get the rights to the publishing from the songwriters. And you would get that by getting mechanical reproduction rights from the music publishers. And in America, you would go to Harry Fox usually to get those mechanical reproduction rights. If you weren't engaged in a different part of the business, if you had a radio station or a TV station, or you were using music in your bar or restaurant, you didn't go to Harry

Fox. You didn't mechanically reproduce the compositions, but you were performing the compositions. So you went to the performance rights societies like ASCAP or BMI. And everyone understood those were the rules of the road. And it was similar in Europe back in those days, there would be a performance rights society and a mechanical rights society. But when digital came up, everyone said, all right, in the beginning, it's downloads. So is a download a performance or is it a mechanical?

And in the US, we had all sorts of controversy about it. And the performance societies were saying they should be involved that there was not only a mechanical, but also a performance because of the download was transmitted to the consumer. This took a long time to resolve. And there were a lot of problems as a result. In Europe, people took what I think is a saner approach. The two types of societies, the mechanical and the performance said, why should people have to deal with this complexity? Why don't we make it easy to license? And we will have a joint venture between our two societies and issue a joint license that covers whatever mechanical and/or performance rights are involved.

And when they got their license fee, they sort of behind the curtain divided it between the performance and the mechanical. But that was something that, you know, the companies that were doing the deals didn't have to really worry about. They got all the rights they needed. America never did that. So then we, a few years later, we go to streaming, excuse me, and people are thinking, oh, well, streaming is like performance. Isn't that like radio where you listen to it, but you don't keep a copy? Isn't that the same thing? And in some ways, it's the same thing, but it's digital. It's not the old-fashioned radio. And when you do stream, guess what? You make a copy in the process of transmitting the music so that it can be played for the consumer. And so the mechanical people said, hey, we have to be involved in this. And, you know, they dug their heels in it. In America, they still to this day get part of the revenue from that. So, you know, it's a combination of

mechanical and performance. So, you know, we have sort of a dual licensing structure here for that activity. And then you go into other parts of the business and it just gets more and more complicated. And you always have to keep in mind who you have to license from. I mean, Europe is complicated in its own way, especially on the publishing side. But in the U.S., excuse me, that's one of the things that gives rise to the most complication. It's fascinating to speak about this. And, Jeff, I'm looking very much forward to hearing you talk about some certain aspects of this on the conference. But anyone who wants to meet Jeff, that's probably a rich opportunity to say hi at the conference. And Jeff, I'm so happy you came on the podcast. I've been following you ever since last year. And I'm honored and I'm looking forward to seeing you in a few weeks.

No, it's great to be with you, Jacob. And I really enjoyed the conversation. And I look forward to seeing you and everyone else at Music Tech Bergen. Thank you.

Thank you.

More from Sound Connections

All episodes