This is Bloomberg Law with June Brosso from Bloomberg Radio.
The Federal Trade Commission and seventeen states are suing Amazon over allegations the e commerce giant abuses its position in the marketplace to inflate prices on and off its platform, overcharged sellers, and stifle competition. The lawsuit is one of the most significant legal challenges brought against Amazon in its nearly thirty year history, accusing it of monopolizing online marketplace
services by degrading quality for shoppers and overcharging sellers. The case also represents a career defining moment for FTC Chair Lena Khan, whose long had Amazon in her sights.
So this case is entirely pro business. It is tens of thousands of businesses that are dependent on Amazon to reach shoppers that increasingly are paying one out of every two two dollars, as well as being subjected to all sorts of arbitrary tactics. So we believe that this lawsuit of where successful will actually entirely restore the promise of
free competition. Our free enterprise system is one where a companies should be competing on the merits and not be able to protect their monopoly power through illegal tactics.
Joining me is Jennifer Ree, Bloomberg Intelligence Senior litigation analyst, jen this lawsuit comes as no surprise.
Oh not at all.
I mean, we've all been expecting this for years. The investigation of Amazon actually started during Trump's FTC and it was already ongoing when the chair, the current chair, Lina Kahan, took her position, which was back in twenty twenty one. And when she did, I mean, it was sort of widely thought that she was hired partially because of her
very overt outspoken antagonism toward Amazon. She had written a long article in twenty seventeen while in law school or at least for the Yale Law Journal, basically saying that she thought that Amazon behaved as an anti competitive monopoly.
So this is no surprise at all.
So tell us what the FTC and the seventeen states are accusing Amazon of.
You know, they're really focusing on it's marketplace, right, because Amazon has other businesses outside of what we as consumers know of Amazon, So they're focusing on the way Amazon treats sellers. And they're essentially saying that some of the policies and some of the conditions they impose on these sellers.
One cause prices outside of Amazon to go up. In other words, sellers if they discount a product where they're selling somewhere outside of Amazon more so than they have on Amazon, haven't offered Amazon the lowest price that they offer, that that seller then gets punished. That's one thing. And they also allege that Amazon punishes sellers by sort of pushing them down in search results or taking them out of the buybox if they don't use Amazon's fulfillment services.
That's using Amazon to store the product package of the product and send the product when somebody buys it. So it's really all about sort of unfair treatment of the sellers that are on the marketplace.
And it's really important where you are in those Amazon on searches, because a lot of people don't get past the first two or three. Lena Khan said Amazon is a monopolist and is exploiting its monopolies in ways that leave shoppers and sellers paying more for worse service. But are shoppers paying more because I find that the prices on Amazon are often cheaper than they are elsewhere.
Well, of course, that's exactly what Amazon's defense is going to be no. We offer convenience and speed and low prices. And part of the reason that we have some of our policies in place that require these sellers on our platform to provide the lowest price on Amazon is because that's what our consumers expect.
That's what we've.
Held ourselves out to be the lowest price that you can get out there in the marketplace.
And this is pro.
Competitive, not anti competitive. But on the other side, you know, there may be some evidence that because it's a little more expensive for a company to sell a product on Amazon than some other website, let's say Etsy or their own proprietary website, that they end up raising the price outside of Amazon on those websites because if they have to offer the lowest price on Amazon, they don't want to kill their margins selling on Amazon. So instead of
lowering their price everywhere, they're increasing it outside. So people who bought that product outside of Amazon are actually paying a higher price. And I think that's part of the allegation of the increased prices.
Amazon is a monopolist. Is that accepted that it's a monopolist or not?
You know, yes and no.
The thing about it address law is whether or not you're a monopolist all depends on how the market's defined. You have to look at the contours of the market. What is the market you're competing in now?
I think the FTC in this lawsuit's basically.
Provided two market definitions. One is the consumer side. It's a broad online marketplace to buy all sorts of goods, and the other one it's a broad online marketplace for sellers to sell all their goods. Two pieces of the market. And I suppose if that is accepted as the relevant market by the court, then you might be able to argue that they're a monopolist, that they have monopoly power. But the fact of the matter is that I at
least haven't yet seeing it defined that way. I've seen market definitions like an online seller of ebooks right that they have monopoly power in ebook sales, or they have monopoly power in certain segments. Because for each of these items that Amazon sells, there are alternatives, and those alternatives may be different depending on what type of item it is. They may be broader, they may be narrower, depending on
the items. So it's going to depend on how that market's defined, and that's something the FTC is actually going to have to support and prove in this trial the proper definition of the market.
I mean, if you had to be on one side of this case, would you rather be on the FTC side or would you rather be on Amazon side?
You know, I think I'd rather be on Amazon side. And the reason I say that is because I think this is a tough case for the FTC to win to start, And I'm not saying that they can't because facts matter, and we don't know what the facts and the evidence are yet. It will matter what they prove in court, what the experts say, and what the testimony is,
and what Amazon's documents look like. But I think the reason I ultimately come down on Amazon side is because I don't think if liabilities proven that ultimately those remedies are going to be particularly drastic. I don't really see a structural breakup. What I think will happen are behavioral changes.
Amazon.
You simply can't do this anymore. You can't push a seller way down in the search results simply because they don't use your fulfillment surfaces, or you can't force them to provide the lowest price on Amazon dot Com, and very Interestingly, Amazon's already promised to do most of that in the UK and in the EU, so it clearly is willing to change some of its rules and change some of the way it treats sellers in order to make these lawsuits go away. But the FDC didn't accept those concessions.
But the FDC in this suit, unlike even mentioning it in the Google suit, is not seeking a breakup of Amazon.
Well, we don't really know yet.
I'd say that they haven't specifically said that, but their language is vague. You know, they're asking for a permanent injunction and they're asking for the court to do what it needs to do to stop the monopolistic conduct. And really what that means is it gives them leeway down the road after liability is proven, because that's obviously the first step to actually seek a remedy, which is a breakup.
And normally when these kinds of cases are brought by the DOJ or by the FDC, they don't explicitly state what they're looking for in a remedy early on. That comes later, and I think they still could ask for something like that.
So Amazon's general counsel said, if the FTC gets its way, the result would be fewer products to choose from, higher prices, slower deliveries for consumers, and reduced options for small businesses, the opposite of what antitrust law is designed to do. And that's what I'm saying about. How can you beat Amazon for like getting it to you the next day?
Well, you know this is why I say, I think this is an uphill climb for the FDC. It's really difficult to argue against the pro competitive aspects of the company. I mean, it has created a very efficient, very consumer friendly marketplace essentially that consumers love. They do get low prices, they do get speed delivery. They makes it very easy. There's one stop shopping, this is all and returns and returns exactly and a lot of help by the way, when there's an issue with the product.
Because I've done that myself.
So I think that when you look at monopolization cases, they're based.
On a reasonableness standard. So what a judge has to do.
Is they have to weigh sort of the harm against the pro competitive side, and whichever side wins out, that's where you land whether you violate the law or not. Is it a reasonable or unreasonable restraint of trade that we're looking at. And I think that when you have strong pro competitive justifications for what you're doing, it makes it harder for a plaintiff, even the FTC, to win a case. And so I think you just put your finger on the head of what the issue is here.
Is this the third time that the FTC is suing Amazon this year or recently.
Something like that.
The other suits are all consumer protection suits, so this is their first antitrust suit against Amazon. The others had to were on their consumer protection side. Quite differ from this.
Fewer states joined the Amazon suit than the Justice Department suit against Google or the FTC's earlier suit against Meta. Is there a reason for that?
You know, It's very difficult to say.
The states probably were all asked, and they weigh the pros and cons of being part of a suit. They tend to be more political the state attorney's general, let's say, than the FTC, I think, and some of them may be more concerned about consumer perception here that consumers tend to really like Amazon, whereas I think the perception about Google is maybe less positive generally when you look at consumer polls, and they also do see the pro competitive
side of Amazon's business. So I think that's probably why you see fewer states joining here than in the Google case, but they could still join going forward.
You know, it's been in business thirty years. Do you think the Amazon sees this as a real threat to its business?
You know, I think that anytime you're sued by the Federal Trade Commission, it's a risk, right and your documents, your information, your executives are exposed what's mostly public trial, it's always a risk, and it can bring bad pr to the company. You know, there certainly have to be somewhat unhappy about it, but I don't think they view it as a long term major risk to the company.
Talking about Lena Kahan, and they have tried to get her recused from the cases against them, right, Yes, didn't work.
You know, it didn't work.
So they got tried to get her recused because it was clear what her position was about Amazon back in twenty seventeen. And that's why we've all expected this lawsuit for so long, because we knew that her view was that this company behaves illegally in an anti competitive manner. The issue is that she's not the one who's going to make this decision. This is in federal court, so she's acting as a prosecutor, and when you act as a prosecutor, that is what you do.
You know, if she.
Did a new investigation of Amazon looking at the same facts that she looked at three years ago, I don't think she'd come out any Differently. Where a lawsuit is brought internally at the FTC, that can be a different matter because ultimately the Commissioners are the appellate panel for that, an administrative law judge makes the first decision and.
Then the commissioners.
The appeal goes to the commissioners and they are acting as judge. And in that kind of an instance, I think it would have been different with respect to recusal. But in this instance, where it's out of her hands now she's just prosecuting, presenting the facts, presenting evidence. Some other party. The judge in this case will make the decision. I think it's less impactful.
Earlier this year, the FTC challenged Meta acquiring the virtual reality company Within and lost that. FTC also lost a similar suit attempting to block Microsoft's acquisition of Activision. So how much will this case define Lena Khan's career?
You know, I think those two cases are quite different from this one, different standards. They were pursued under a different antitrust statute, and to be fair, by the way, the FTC is still appealing the Microsoft activision decision, so it's not completely finished yet, even though I do think
the companies will be able to close that deal. I think these more than those, this one, this Amazon case more so than those two, because it's just a long time coming, and it's been her goal to try to pull the way antitrust laws have been interpreted back to the way they were interpreted more like in the nineteen sixties. You know, there was a big change in the nineteen seventies and nineteen eighties and the way the antitrust laws were interpreted and then in the way that enforcement played out.
And her view is that because of that change, anti trust enforcement became too lenient and too lax, and that we have to go back to where we were sometime before the nineteen seventies, where we really looked at market structure. It just made a presumption that if a market was concentrated, it was likely to cause harm to consumers, rather than looking at whether prices to consumers are going up or
output is getting low, but putting the structure aside. So even if it's an oligopolistic structure, so long as prices to consumers are low and output stays up, we're okay. And she's trying to bring it back to where it was. And I think this suit is a big step in that direction, as those two merger suits were, but they were sort of baby steps.
This is a big step and we'll see just how this big step goes for the FTC. Thanks so much, Jen, that's Bloomberg Intelligence Senior Litigation analyst, Jennifer Ree.
I will be working alongside humans to provide assistance and support and will not be replacing any existing jobs. Sure about that, Gus, Yes, I am sure.
It was the first human robot press conference. In July, a United Nations tech agency assembled a group of robots that look like humans to answer reporters questions about the future of artificial intelligence.
I think my great moment will be when people realize that robots like me can be you to help improve our lives and make the world a better place. I believe it's only a matter of time before we see thousands of rombots just like me out there making a difference.
That time is already here in many respects. Apple's Siri has been responding to your questions for more than a decade, and the release last year of chat GPT has opened a worldwide debate about artificial intelligence and led to the
filing of lawsuits over intellectual property rights. The latest a proposed class action by more than a dozen well known authors, including John Grisham and George R. R. Martin, against open Ai for copyright infringement, calling its chat GPT program a massive commercial enterprise that relies on systematic theft on a mass scale. My guest is intellectual property litigator Terence Ross, a partner at Katin Yuchen Rosenmann. This is, I believe,
the third lawsuit like this over chat GPT. What's about the complaint?
So, this particular lawsuit is brought by the Author's Guild, which is association that represents authors for various reasons, and here they have taken the position that open AI's chat GPT, which is an artificial intelligence system with a learning module built in so that it can actually improve its ability
to function by learning. But the Author's Guild has alleged that feeding the chat GPT the works of its authors to help it learn, and then the chat GPT using that to answer queries, which often involves quoting passages from
the author's works, is a copyright infringement. And the lawsuit is brought as a class action on behalf of a very large class of authors here in the United States, and it'll have to be determined at some subsequent date during the legation whether or not it's a legitimate class action or not. But it's a sally all the authors in the United States suing open Ai, the owner of chat GPT, over how they're using the author's works through chat GPT.
In another suit brought by authors last month, open Ai move to dismiss the complaint and argued that the training basically constitutes fair use. So it seems like their defense is going to be fair use.
That's absolutely correct, and a fair use's statutory provision in the Copyright Act of nineteen seventy six that provides that in certain instances, copyright work can be used for secondary purposes that society deems as worthwhile useful. You know, the most important from your perspective is for news gathering and news broadcasting. But teaching is one of the expressly listed peeps of uses secondary uses that would be considered fair use.
And it's a very interesting defense. Now, the statute Section one US the Copyright Act actually says nonprofit educational purposes, and that may pose a problem to open AI because I'm not sure if this would legitimately qualifies a nonprofit educational purpose or not. But it's a colorable defense.
Quite Frankly, we talked last term about the Supreme Court's decision in the Warhol case that rained in the scope of fair use. So do you think that courts are likely to maybe rein in fair use in this case?
So I'm not sure that the Warhol case is going to have any impact on this case whatsoever. I think the Warhol case does reign in the uses of the defense, the fair use on the margins. But as I understand the defense being raised by open AI in these lawsuits, it's really the core of fair use. You know, the statute was intended to protect certain types of secondary uses,
and it listed examples of them. It's not all encompassing, it's not exclusive list, but one of the expressly listed purposes and nonprofit educational purposes, and therefore this is at the core of the fair use doctrine, and I really don't see how the Andy Warhol case will limit that anyway.
The authors claim that chat gpt can produce works that mimic their books, and there are businesses that sell prompts allowing users to create what's essentially works of fan fiction.
That's exactly what the concern here is. And in one sense, the fair use defense being raised by OpenAI is really diversionary. They are saying that they are feeding all of this data into chat gpt so that it learns, and part of that process of feeding data into it is feeding into the machine entire novel as well as fact based works which have lower copyright protection. That's all being fed
in purportedly so that the machine learned. The problem is less with that, in my view, than what happens after it learns. It's the output of chat GPT that is
the problem. And when it, in response to queries quotes copyright it works, or on behalf of a query from an individual, creates a purportedly original work that incorporates copyrighted language or large portion of a copyright work, that copyright infringement in the classic sense straight up and up copying, and that is the core problem here, that there is no apparent restraint on chat gpt from doing that in response to queries from individual users of the machine.
And there's also a question of whether you can get a copyright on works created by artificial intelligence, and there was an interesting first time ruling on that by DC Federal Judge Beryl Howell tell us about that.
So the whole area of artificial intelligence is raising a host of problems for intellectual property laws. The Author's guild lawsuit against Open AI is one aspect of that. Another aspect is whether the creations of chat GPT or other AI type machine is protected by intellectual property. And so a gentleman by the name of Stephen Taller has an AI system that he refers to as the Creativity Machine, and it created, according to him, an autonomously generated piece
of visual art. I eat, a painting which the machine entitled a recent entrance to paratus. It's actually quite attractive painting. And mister Taller applied for copyright registration with the United States Copyright Office on behalf of the Creativity Machine, reporting to the Copyright Office that the painting had been autonomously generated by his AI machine, and the Copyright Office rejected that application on the grounds that copyright only protects creations
by human beings. Mister Toller appealed that ruling to the United States District Court for the District Columbia, where the Copyright Office is based, and that court just issued a very important opinion of first impression saying that artificial intelligence is not entitled to claim copyright in anything it quote unquote creates, going further to say that human authorship is
an essential, unrequired part for a valid copyright claim. It's a very important decision in this field of artificial intelligent will almost certainly be appealed to the DC Circuit.
I mean, was the judge on solid ground.
So Judge Howell's decision is actually really quite good. The DC courts do not often see copyright cases. It's just simple fact. We've talked about this before. The bulk of copyright cases come from the Second Circuit of New York in the Ninth Circuit California, specifically Los Angeles. And so this was usual and so really outstanding decision by Judge how in which she lays out the history of copyright
in the United States. So she points out that James Madison in the Federalist papers referred to authors as beating persons. And that's important because after we actually got the government up and running, he was a congressman in the House Representatives, he was on the committee that drafted the first Copyright Act. That Copyright Act, very first Act uses words like executor, administrator, administrate, takes he, she, which seems to imply that copyright is
held by humans. The nineteen oh nine Copyright Act expressly described the right a quartered by copyright as going to a person. The nineteen seventy six Act, the one we currently operate under, has multiple references to people. For example, Section two three the Copyright Act sss when an author is dead, and that's a quote that's implying that the author has to be a human. It uses the terms widow, widow, we'er, surviving children in connection with succession of ownership of copyright.
These are all indications that Sugg points to in the history of copyright that indicate that copyright is limited to human beings. Of course, we had that famous case a couple of years ago I have I think it was a ninth circuit, the Rudo versus Slater about a monkey selfie. A monkey had somehow I guess accidentally triggered a camera and taken a picture of himself. And the ninth Circuits that know that doesn't work. That human beings have to
be people who get copyright. So there is this case law as well as the text of the language of the statue which supports this notion that it has to be you would be something. You know, we've had this long history of copyright having to deal with new technology.
When fortography first came around, there was lots of questions about whether a photograph was copyrightable, and the Supreme Court said, yeah, it is because there's a human who's involved in controlling the process and making decisions like lighting poses, you know, how to develop it. And that was sufficient human involvement to justify photographs being copyrightable. And more reason, they's computer programming.
The human being writes something called source code, computer then translates that into object code, which is quote unquote machine readable code, so the computer can actually process the bits and bytes you know, one zero zero one one zero. And the supports have said, well, that again involves human development and control, and the computer is merely translating the human actions into machine readable code, and so that's copyrightable.
But now we're at a point where we're saying, at least in this case, that there was no human interaction, that this painting was a comanimously generated by an AI machine. And Judge Hall makes an important point that human activity is required even when technology is used to some extent. Praises a lot of questions for a future. But I think there's a decision score easily hold up on appeal Terry.
I'm curious. So when Taylor made the application of the Copyright Office, was it in his name? He was asking for a copyright for himself.
No, he was asking for a copyright on behalf of his creativity machine. He was asking for a copyright on behalf of the AI. Now, interesting thing happened after this case came out of the Copyright Office and was taken to the District Court in DC. All of a sudden, mister Taller started talking about well he issued prompts to the AI machine to do this, and Judge Howell said, well,
that's not the record before me. The record before me is you reported to the Copyright Office that the AI autonomously generated this painting, and under that set of facts. There's no copyright issue because machines can't get copyrights on
the humans can. But it leaves open this question, and one of my colleagues here kat has argued that there may be a point at which there are so many human prompts to the AI that the resulting work is copyrightable because it fits into this notion in the history of copyright that some human activity and control is required. Now, the question is how much of that is going to be acquired in context of ag on a few prompts.
Probably isn't going to be good enough. So talking about the author's guilt case, if some high school student says, write an essay for me, and the machine writes the essay, is that sufficient of a human propt? Probably not, And to the extent that it then just regurgitates something from a copyrighted novel that's out there, such as John Grisham's work, that would indeed be copyright infringement on the part of
both the student and the AI machine. So I don't know where the courts are going to draw that line, or whether they will at all, Whether the to say no AI is never going to be entitled to copyright the registration or at some point they say, well, if you give it a thousand commands, maybe that's sufficient to get you. You, not the AI, but the individual who gave the commands a copyright registration.
Had there been a lot of other federal court judges ruling on AI and copyright.
This is the first of its kind. We lawyers call it a case of first impression. And why it's so unusual that it's here in the District of Columbia instead of New York or Central District, California or Los Angeles is that the Copyright Office is located here in DC, and so you take appeals from their decisions to the DC District Court and it work itself up to the DC Circuit Court, which gets a copyright case maybe every
other year. So again they'll be writing on blank slate, and I assume this will go all the way up to the Supreme Court eventually, and we'll see what they say.
Coming up next, I'll continue this conversation with Terrence Ross and we'll talk about patents and artificial intelligence. I'm Jim Gross, and you're listening to Bloomberg. I've been talking to Terrence Ross of Katin Yuchen Rosaman about intellectual property and artificial intelligence. This is copyright? How is patent law handling AI?
The issues of artificial intelligence are not limited to the copyright realm. The same gentleman whose AI created work was involved here, Stephen Tholer, had previously applied for a patent with the United States Patent and Trademark Office on what he claimed was an invention developed by an artificial intelligence, and the Patent Office follows similar route to the Copyright Office and said that no, machines can't be inventors for purposes of the patent laws, only humans can and rejected
the patent application. He took that on appeal to the United States Cord of Appeals for the Federal Circuit, which is where you appealed decisions of the the Patent Office, and the Federal Circuit agreed on with the United States Patent Office and said only humans can be inventors for purposes of patents and affirm the decision the Patent Office.
Mister Poler, apparently having unlimited resources, then filed a petition for rit asserts sory on that decision with the United States Supreme Court, which just this past spring denied that petition. So the law is on the patent side the same as on copyright side, that humans are the only ones who can be listed as inventors on patents. And he
goes back for in a time. Every now and then corporation supply make the mistake of applying for a patent in the corporate name, and the patent offices no, no, no, no. The only the inventor, the human can. They can then assign the patent to the corporation. But corporations can't be inventors. Only individuals can. So it's consistent with past patent office practice,
it's consistent with the approach taking and copy right. So we're starting to see this evolution within intellectual property that comes to the same point, which is that only humans are entitled to intellectual property rights in the United States.
What about these robots that have the characteristics of humans? Can you get a patent on those?
Inventors can get patents on an artificial intelligence system, just as they can get a patent on most other computer software. AI is usually more than just software. It's usually some sort of system, so there may be a method patent as well as utility patent on it. But no, I'm actually representing someone right now who has obtained a patent in the AI field. That the really interesting question, though,
is this we're using. We're seeing the use of artificial intelligence in a lot of activities that traditionally are considered creative. One example is encoding writing software programs. There are now artificial intelligence programs that help programmers to write code. You set up a framework, you say I want to do this in this section of the program, and the AI
writes that code. Similarly, in Hollywood, we see movies and television artificial intelligence doing special effects and providing and being involved in some elements of a production. Indeed, the recent writers strike and Actors strike had at their core demand by the unions that AI be prohibited in movies intelevision. But in both these contexts and the use of AI to help encoding, the use of AI to help in
film and television. Now that we know that AI generated work is not copyrightable, what impact does that legally have upon the computer programs, the motion pictures, the television shows that have used artificial intelligence as part of the creative process. All of a sudden is Star Wars deprived of its copyright because artificial intelligence machine helped on portions of it is a software program or a video game, which is
essentially software program. Are they because they have used AI to help in the coding, have they lost their copyright? I mean, this is a really important legal issue that a lot of companies rushing to use AI have not considered the questions whether in the act of employing A help in the creation of your work, you're losing the ability to obtain copyrights. I mean, this is for studios. This should be something that's on the front page of their agenda of things to think about. This has to
be on their radar screen. Have never understood why they were so opposed to the union demands on artificial intelligence, because in a fact, by incorporating AI, they're taking away the most valuable asset they have, their copyrights in their motion picture and it's something that's you don't hear much discussion about.
Maybe that will be the next topic we hear a lot about. Thanks so much, Terry. That's Terrence Ross of Catain Euchen Rosenman, And that's it for this edition of the Bloomberg Law Show. Remember you can always get the latest legal news on our Bloomberg Law Podcast. You can find them on Apple Podcasts, Spotify, and at www dot bloomberg dot com, slash podcast, Slash Law, and remember to tune into The Bloomberg Law Show every weeknight at ten
pm Wall Street Time. I'm June Grosso and you're listening to Bloomberg