Transcript: Article 13: An Unlucky Number for YouTube

Recorded at Copyright & Technology 2019

with
• Christiaan Alberdingk Thijm
• Tom Frederiske
• Jean-Baptiste Soufron

For podcast release January 21, 2019

KENNEALLY: We live in divided times. In the United States red state senators oppose every budget bill from their blue state counterparts. In the UK, leavers argue with remainers. In Paris on the weekends the air is shrouded in tear gas, and the Arc de Triomphe laced with graffiti. Usually the sides are evenly matched. Any victory achieved won’t be enjoyed for very long. Tomorrow we will see another vote in Parliament, another election, another demonstration. I wonder whether the politicians think they’re kidding when they propose a second Brexit referendum. Take this as a safe bet, another vote is only setting us up for best two out of three.

The world of intellectual property law is no less divided than anywhere else we look. In 2019, and in fact, since the dawn of the digital age, the copy left are constantly tilting at the forces of copyright. Information wants to be free. On the other hand, information wants to be expensive.

Yet in Brussels last summer something remarkable happened. European Parliament agreed overwhelmingly, though, yes, after a very long period of legislative horse trading, to the text of a long-awaited directive on copyright in the digital single market. It’s not a done deal, not quite yet, though it is inches from the goal. A final vote could even happen this week. And if everyone isn’t happy, nearly everyone is getting ready to live with it. Be careful what you wish for, of course.

Several articles in the directive have received special attention. Article 13 could require certain types of online services to take proactive steps to license copyrighted material or else keep it off their services. It will take years, however, before we learn what types of technologies or enforcement schemes may be acceptable or required under the law as implemented in EU member states.

Regardless, just as the EU GDPR, that’s the General Directive on – I’m sorry, General Data Protection Regulation had implications for websites terms of use worldwide. The new EU directive could have a significant ripple effect on the copyright-related responsibilities of online services here in the United States and beyond.

To discuss the directive and get a head start on predicting its implications, I want to welcome my panel. We’ll start from the very far left, and bienvenue to Jean-Baptiste Soufron. Jean-Baptiste Soufron is an attorney at law at the Paris bar, specializing in media, intellectual property, and digital law. He was formerly general secretary of the French National Digital Council, as well as a senior advisor for digital economy in the cabinet of the French Minister of Economy and Finance. His other previous positions include director at the think tank of Cap Digital, and he was chief legal officer of the Wikimedia Foundation.

To Jean-Baptiste’s right is Chris Alberdingk Thijm. Chris, welcome. He joins us today from Amsterdam in the Netherlands. He’s a partner there at Bureau Brandeis. He specializes in digital copyright, the liability of intermediaries’ privacy and free speech issues. Besides his work as an attorney, he’s a senior lecturer at the University of Amsterdam, as well as the author of a novel, The Trial of the Century, which was published in 2011 and is currently in its seventh edition.

And then finally, immediately to my left is Tom Frederikse, joining us, joining us from London today. Tom Frederikse is a partner at Clintons solicitors, where he specializes in digital media and technology related issues. His extensive practical experience in media and technology allows him to work on a wide range of commercial media matters, including content licensing, music publishing, PROs, trademarks, and data protection. He is dual qualified as both a UK solicitor and a New York attorney, so we might get two perspectives there for the price of one. And previously Tom spent 15 years as a music producer and engineer, so a fine musical background to add to that. Again, welcome to the panel.

Chris Thijm, I want to open with you because you’ve offered very kindly to help familiarize everyone with the important provisions of Article 13. Some may be familiar, some may not, but I think we need to set some groundwork here. Can you run through that in brief order?

THIJM: Sure, sure. First of all, can I see a show of hands, who of you are European? So we see a few Europeans.

KENNEALLY: So we see a handful.

M: (inaudible).

THIJM: Some Europeans, then. Well, it’s really tremendous, it’s really fantastic to have so many Americans in room discussing a draft revision of a European copyright directive. It’s really amazing. I said yesterday while we were having dinner, I can’t really imagine us in Europe discussing the new (inaudible) act, but it really shows that this is a very controversial provision, and there’s a likelihood that it may be export, but I’ll get to that.

But first of all is Article 13, and a lucky number. Well, it very much depends on who you talk to. Some say, and the ones who’ve been lobbying for this provision say, that this is the provision that covers or rescues or prevents the value gap. What is a value gap? The term was coined by lobbyists from the copyright industry in Brussels. It basically means that there’s a gap between what YouTube makes and what the creators, at the end of the day, receive as remuneration for their works being exploited on YouTube. In Paul Goldstein’s analogy, YouTube is the celestial jukebox, but nobody’s getting paid besides YouTube for the use of the celestial jukebox. So it’s free. Others call Article 13 a Google tax or a YouTube tax, but that’s not highly original because all new legislation in Europe is called a Google tax nowadays, I’m afraid. We are not very fond of Google in Europe for some reason. It is also dubbed a meme ban, and others say that the provision will lead to internet filtering.

There’s a lot of protest against the provision. YouTube recently started a big campaign in November. If you go online you can see people from all European countries, so from all 28 member states, arguing, especially to youngsters, why the provision is bad. So it very much feels a little bit like the Pope, (sp?) PIPA, and SOPA legislation. The proposals you had here in the US a few years ago, there was a little backlash against, and the internet blackout, and what have you. It’s a little similar here. And we’ve had Tim Berners-Lee and Jimmy Wales, and other important people that have stood at the foundation of the internet speak out against the provision.

So why is it controversial? Well, to understand that, we have to go back a little bit in time to the year 2000/2001 when, in Europe, on the one hand, the so-called e-Commerce Directive was formed, and the latest copyright directives, the big copyright directive from 2001, the internet or the InfoSoc Directive. In the e-Commerce Directive, the e-Commerce Directive was drafted with the example of the DMCA in mind, and the Safe Harbor provisions that you have here in mind.

So the e-commerce directive grants certain types of activities, not types of services, but for certain types of activities, an exemption of liability, in this case, importantly, for hosting. So if I am hosting content of third parties that they’ve uploaded, I, as a rule, am not liable for that content unless I’ve been notified of the unlawful nature of the content and have refused to take it down. So it’s a basic notice and takedown system like you have in the US. It’s a little less sophisticated, but that’s the way it works. As a result, of course, these hosting providers are not violating copyright by allowing others to upload copyrighted content to their platform. In short, YouTube currently is not liable for the content on its platform, nor is it performing copyright-protected act. So it’s not performing a communication to the public.

And this is what is going to change under Article 13. Article 13, and you have it before you, and I saw that the Article, too – there’s a part of the sentence missing, it fell off the page, probably, but we’ll get to that. But you have it before you. It says in 13, paragraph one, that online content sharing service providers perform an act of communication to the public. So this is very important because it means that by allowing others to upload content to my platform, content that I may have no knowledge of whatsoever, the platform is committing an infringement upon copyright. As a resulted (sp?) of the second sentence says, they shall therefore conclude fair and appropriate licensing agreements with rightholders. So they have to conclude licensing agreements because they are directly liable for copyright infringements. So not only are they losing their exemption from liability, not only are they losing their Safe Harbor protection, they are also violating copyright law directly. Now this may, at the end of the day, force YouTube – and this is what YouTube is arguing – to drastically change its business model. YouTube is basically arguing we will have to ensure that no content that isn’t clear beforehand will be uploaded to our platform. That’s what YouTube is arguing in this big campaign.

Now why is this important for US? Why is this important for you? I think we can mention a number of reasons why this is important, but I’ll mention three here.

One, I think it’s important to note that copyright laws get exported. We have seen Paul Goldstein this morning discussing the term extension here in the US in 1998 from life plus 50 to life plus 70. Well, that was a European invention. We had it five years before, and then it came to the US.

Second, copyright law has a tendency to harmonize internationally. Who remembers what happened in 1996? So we had OJ Simpson in 1996, Monica Lewinsky, 1996, but also the WIPO Copyright treaty in 1996. WIPO Copyright Treaty that harmonized copyright laws internationally. And that will also happen here, I may fear.

Thirdly why it is important is that in international copyright law we have this rule called the Lex (inaudible) meaning that the law will apply of the country where I seek my protection. So if I’m in the Netherlands and somebody in the US has uploaded my protected content on YouTube, but it violates my rights in the Netherlands, I can sue that person directly in the Netherlands for copyright violations for an act that is in itself not contrary to US copyright law. So that may also be a big issue.

There are many more point to address or issues to make in relation to this directive, but I have two very knowledgeable panelists here also, and I don’t want to speak too much, but we may be getting back to text of the provision.

Oh, yeah, one last thing that is important to note – thanks, Chris – is that should you not be able to conclude – and this is where the filtering bit comes from – should you not be able to conclude licensing agreements with the rightholders because the rightholders don’t feel like it or they’re asking an outrageous price, then you have to take measures in order to ensure that infringing content is not uploaded on the platform. So many dub this as the filtering provision ensuring that measures have to be in place in order to prevent content from being upload. And YouTube is arguing – well, (inaudible) for a video content platform and in video, the likelihood is that there are a multitude of rightholders concerned, it’s going to be very difficult for us to clear or ensure that the people uploading the content have cleared the content at all because the multitude of rightholders in relation to video content may lead – this is what YouTube is arguing – to them having to block content altogether.

KENNEALLY: I really appreciate that very thoughtful and high level overview. We’ll continue at a high level for a moment before we get into some of the specifics around it. But I do want to call out something that people may not entirely be aware of. With the GDPR, that became effective across Europe on a date certain last May – May 25th, if I recall correctly. That’s not the case with this particular directive. But there’s an important difference here between directive and a regulation. And so this matters to this American audience as much as it matters to the Europeans, so explain that if would.

THIJM: So we have two – well, we have a lot of legal instruments, but two main legal instruments. One is called a directive and the other is called a regulation. So the GDPR is a regulation, and that means it’s a law that will have immediate effect on the date on which it enters into force as a law throughout European Union, so in all 28 member states. The directive is different because a directive needs to be implemented within the national laws. So that means that in this case there will be a grace period of two years for the member states in order to implement the new copyright directive into the national laws, and it also gives the member states a little leeway to deviate to a certain extent. Of course the directive is intended to harmonize copyright law in Europe, so it’s not intended to have member states have completely different copyright laws. But you will see some details differing in various member states.

KENNEALLY: And for something like YouTube, those could be expensive details.

THIJM: Yeah.

KENNEALLY: Absolutely. Well, let’s turn to Jean-Baptiste Soufron, because you have an interesting perspective on Article 13 and what its intentions are, and how much you like the way it’s written, as well. And so let’s open with the thought that what we’re really talking about here, as much as it labeled a copyright directive, is about a European perspective on antitrust law. Why is that how you see it?

SOUFRON: Yeah, so this is something that is becoming more and more common with EU regulations lately. Meaning that as it is extremely difficult to use antitrust law online, it doesn’t work easily. Well, they began using a lot of other regulations to achieve antitrust means and objectives. It’s what happened with the GDPR, which contains a lot of antitrust objectives in it and antitrust regulations like for example the obligation to allow your customers to transfer their data from one platform to another – that’s typically an antitrust solution. And in here, when we look at it, it’s not written with the words of copyright law. It doesn’t look like copyright law. If you put in the French intellectual property code, it will certainly stand out as something really different with its own mechanism and logic.

The other thing, instead of talking about it only as if it was a new copyright mechanism, we should understand that this text is not – well, it’s actually an antitrust text. It will be the first time, to my knowledge, in copyright law that will have a different kind of legislation applying, and the other criteria, the distinction – well, making the distinction between a small company and a big company. As of today, if I’m a photographer, and you use a picture that I took, well, I don’t care if you are a big newspaper or an interviewer. Maybe you will have some exception or something allowing the use. But at the end of the day, it’s the same thing for everyone. Not anymore. Now we’ll have a special text for Google, Facebook, and Twitter. And maybe for the Chinese equivalents that will come in the future.

So it’s something that is more and more common and it’s happening with this text. It’s also happening with upcoming text like probably the new (inaudible) new text will also have a lot of antitrust built into it. It will probably be the same thing with the new EU e-consumer regulations coming in.

So it’s something that is complicating the process, and of course it brings like creating some sort of chimera because all of a sudden, I don’t know how people will react to it, but you have a lot of compliance issues with this text. You have compliance obligations, you have governance obligations. You have a lot of stuff that is new to people, but it’s not that new to antitrust people. We’ll have to learn how to use it, and I think it will be extremely complicated. It will be as complicated for people trying to defend their copyright as for people trying to use that copyright. It will be a difficult text to apply. But well, at least one sure is certain is that I think that this text shows – well, it’s a trend. Just as GDPR, just as other texts have tried to, we are coming probably to the end of the – how do you say that in English, the mere conduit.

KENNEALLY: That’s it.

SOUFRON: The mere conduit.

KENNEALLY: It’s just a mere conduit, yeah.

SOUFRON: A conduit, yeah.

KENNEALLY: Which was – to enlighten people here, that was really how it was summarized on a bumper sticker, the e-Commerce Directive, back in 2000.

SOUFRON: And so we are probably beginning to feel a mere conduit fatigue. Nobody cared about the mere conduit for long because these were startups, and what they were doing was not important. And they needed to develop and get big. And now we have the Brexit, we have Trump election, we have all these platform that are not liable at all for what they do online, and people – yeah, people are fed up with it somehow. So –

KENNEALLY: Right. And what’s interesting about it is these are the same targets in the GDPR as in Article 13, and they coincidentally are often, in fact, exclusively if I think of the three you mentioned – Facebook, Google, and Twitter – American companies.

SOUFRON: Of course.

KENNEALLY: And that, then is a place to hang a further thought regarding the inability in the US for anyone to come to any real consensus around antitrust as it relates to these large internet companies. And your question is, isn’t there anyone here who can write this? It really is as if Europe is writing US legislation by default.

SOUFRON: Yeah, and it’s a problem. Well, it’s a problem and it’s not a problem, because at the end of the day we need to harmonize stuff, so it’s cool. But still we are seeing, with this text, especially with Article 11 and Article 13, we are seeing a text that has been intensively lobbied and cowritten or rewritten by US scholars trained to understand EU law, and it doesn’t work because most of them don’t get the way it work, especially lobbyists, actually. And then they tried – yeah, it’s like for many EU lawyers, this text will be so difficult to apply, understand, transform, and I don’t know if it will be that easy to trans – for the GDPR it’s more easy to actually understand how you can apply it in the US and how you can transfer it to the US or to India or somewhere else.

For this one it will be very difficult, because what’s created there is kind of complicated to use. But still we need to come up with a solution about that because is it normal that the global internet legislation is being written in Europe with everybody fighting, like at the Parliament, at the commission, they should be discussing this – international bodies and organizations just the way it was meant to be and the way it was done before. That’s how you create a good text, especially when you have so many international companies involved. Because I said this is an antitrust text somehow. But if you were – if I take the point of view from Google, for example, they will say it’s a nationalistic text, and it’s just Europe trying to protect its enterprises against US companies.

KENNEALLY: Right. So Tom Frederikse, I want to turn to you in an interesting position because your law firm represents not only rightsholders, and the list of music acts is impressive indeed – the Who since 1960, the Sex Pistols since 1976, Radiohead, Paul McCartney, many, many others. But you also represent some of the streaming services, as well, which – both sides have very different interests in this. Outline for us how the rightsholders feel about Article 13 and how the services do, and where this may end up leading us to the law that’s not in any books, but the one we all know, which is the law of unintended consequences.

FREDERIKSE: From our point of view, rightsholders don’t really care too much about the lawyerly points that have just been raised. It’s true that the law would create a more certain position where YouTube is committing a communication to the public and so forth, but rightsholders have never had any doubt that YouTube is infringing copyright. That’s not news. The news is that they have an immunity, as Christiaan pointed out very clearly, and they’re pissed off about the value gap. I’ll use a slightly different version of that value gap. Most of our record company and artist clients consider the value gap to be the gap between what YouTube pays for a stream and Spotify pays for a stream. So in other words, Spotify is parting with something like 70% of all of its gross revenue and paying out something like a penny per stream, whereas YouTube is parting with about 40% or 50% of its revenue and paying out about a quarter of its stream – a quarter of a penny per stream. So that’s a lot less money that you get from a YouTube stream, and that’s, I think, more of where the value gap is.

So most of these niceties about antitrust and so forth are really not at the forefront of rightsholders view. They think that Article 13 would be just fine. There’s no problem for them there. The worst that could happen, as a result of Article 13 – what’s the worst that could happen to DSPs? Well, they may have to use upload filters or perhaps content checkers if it’s a filter that happens after the upload. Well, what is an upload filter? I’d say it only has one template, and that is YouTube’s Content ID. That’s what an upload filter is because there isn’t another one in the world, so that’s the one that we have to look at. YouTube would be legally required to use upload filters perhaps under what looks like is going to be the final text, and that would mean that they have to continue doing exactly what they’ve been doing for the last few years. The only change that probably would result is that they wouldn’t be able to favor particular rightsholders for commercial reasons. They would have to apply it evenly across the board, which is a fairer way to go about it.

This means all other DSPs would then have to create a system which resembles Content ID, that is if they have $10 million or $20 million around for development costs. If they don’t have that kind of money to create such a sophisticated and difficult tool, complex tool, as Content ID, then they’re going to have to perhaps license one, and that means perhaps Google will go into the business of licensing their Content ID software to all other DSPs which ironically would mean that there’s even less money from these new startups to go to content owners. So that would be an unfortunate result.

The legal view here, though is that no court is ever going to be able to perfect a text standard that says what is an acceptable filter system? So how do we decide what is an acceptable filter system and what isn’t? Well, it’s my suggestion that the only way they could do this is to compare it to the only existing filter system that there is, Content ID. And if you do what that does, then you’re a filter system. If you don’t do what that does, then you’re not a filter system.

The law also has some lip service to the idea of a proportionate basis so that not everybody has to have exactly what YouTube has. But the word proportionate also is very worrying to rightsholders because we all know that YouTube is something like 100, 200, 1000 times bigger than most DSPs. Does that mean that the standard you have to reach is 1% of Content ID? So if I deploy a system that is terrible, but is proportionately equal to what Content ID is being – what Google does with Content ID, would that be good enough? That’s certainly a concern.

Also, the directive uses alternatives to a requirement for a upload filter by saying best efforts or measures or cooperation to avoid infringement. Well, what could that be? It has to be an upload filter. There is nothing else. You can’t have a chat about trying to get your infringing content off of YouTube. You either know that it’s there, or you don’t. So I think it looks like the worst that could happen would be they have to use upload filters, and rightsholders would not have a problem with that.

I would also just say that in the rightsholders eyes, the current notice and takedown regime is not effective. It has become an arms race that is not pleasing to rightsholders. I saw that the top three copyright owners in the whole world in terms of those who send takedown notices to YouTube were recently published, and rather weirdly to me, the number one takedown person in the world is the British record industry, the BPI. I would have thought it would be the US’s RIAA that runs the record industry in this country. It’s not. For some weird reason, Britain has focused more on takedown notices, and Britain has sent 424 million to date. That’s about 100 million per year, or about three per second, 24/7, all year long. And all of those are being responded to by Google expeditiously as they need to. And of course, we all know what happens. Instantly it’s taken down, some kid puts it back up. So it’s just a relentless circle, has gone on for many years now, and this would be replicated clearly in the filters arena. So that is certainly a worry for rightsholders, that this would effectively become an arms race of who can build a better filter system, and then who can combat it.

The only other issue that I think is important in this Article 13 for rightsholders that they focus that we see is notice and stay down, which would be a very, very attractive regime if it could be implemented. The idea here is, and I think this – it doesn’t say this in the legislation, but I can’t imagine any other way that it would work. The idea is that when a takedown notice is sent, a piece of content is taken off of YouTube. At the point when it’s taken down, a snap shot is taken of that content, so that if anybody tried to upload it, then it’s blocked. Therefore it cannot be put back up. That’s notice and stay down. At least it couldn’t be put back up by the same person who had posted it anyway.

But there are three huge issues with this. First one is that the big takedown issuers, like the BPI, who can do it with every – three per second will be more protected because obviously they have the bigger arms and they are able to send those takedowns and to enforce those stay downs. So we’ll end up with a situation where big, rich rightsholders effectively win over small ones. Secondly reuploading would no doubt be gamed. The kids are doing things like speed changes and masking techniques and so forth in order to upload what they want to. Obviously, they’re very good so we have a problem of better mousetrap and better mouse. And then, of course, the one that hasn’t been mentioned yet, I think the rightsholders are cognizant of, but it’s a very difficult subject is if you use these kind of upload filters, then you got the problem with fair use because it’s very, very difficult for an electronics system to have any kind of contextual awareness to figure out when something is fair or has an exception that should apply to it.

So in summary I think here the idea is that there is no silver bullet that’s going to solve the Article 13 problem, the Safe Harbor problem, which has turned into an arms race, with any kind of certainty or proper justice. But this isn’t a bad way to go about it. And many of us lawyers have been staring at the DMCA and Safe Harbor for more than 20 years now, and we’re dumbfounded. We can’t figure out how to solve this problem. Here is a constructive suggestion on how to try to mitigate the effects on rightsholders, and perhaps it’s worth a try.

KENNEALLY: And Tom, I’d like to ask you a question that I’ll then turn to the others to give their response on, as well, which is as I said in my introduction, these kinds of battles take place in very close quarters. Victories are momentary. Sometimes there’s never really an end to it. How did it happen that this summer the vote was so lopsided in favor of the directive entirely, even with the opposition that was gathering momentum around Article 13?

FREDERIKSE: Lobbying is obviously a crazy thing these days. You mentioned the lobby effort against America’s attempt to temper Safe Harbor back in 2010 or ’11, it was. I think what we saw were two crucial moments, one in June when the vote went the wrong way for rightsholders, and in September when it went the right way for rightsholders. And I’m not entirely that those votes were informed by really good information and understanding. I think a lot of it was pure lobbying.

KENNEALLY: And Chris, have you got a view on that? What kind of flip to the Parliament on this?

HIJM: It is a severely lobbied piece of legislation, and it’s important to note that where Jean-Baptiste has been talking about US legislators basically creating European legislation, or through lobbyists, this is what happens all the time, and this happened in 2001 when we had our copyright directive. We are creating IP legislation on the basis of US lobbyists that hope that that legislation will ultimately be exported to the US, as it very often happens. You can see who lobbied well if you read the text. Wikipedia lobbied well because that’s excluded from the scope. eBay did remarkably well because marketplaces are excluded from the scope somewhat surprisingly. You can see that open source software developing platforms did well because they’re also excluded. And then Jean-Baptiste already said the small and medium enterprises are also excluded, which is peculiar because in the old days we were always talking about technology neutral legislation. We wanted to apply the legislation to whichever technology possible, and not make it specifically dependent on the type of business or the type of technology you were offering. Well, this is completely the other way around, so it’s very specific, and as Tom rightfully said, it’s going to be very difficult to pinpoint exactly when the article will be triggered.

In responding to Tom’s comments about an order to stay down, the provision actually goes farther than that, because you’re liable upon the content being uploaded on the platform. So that does put a very heavy burden upon YouTube. The Content ID system has very often been described as a kind of remuneration right system. It was YouTube saying, OK, you want to take it off or you want to receive a remuneration on the basis of advertising. We’re willing to do that. So if you look at that through a copyright lens, the exclusive rights are there, so the idea that you are an exclusive rightsholder and can say, no, you’re not allowed to have that can determine the terms of the remuneration you are being paid, that is more or less left. And this article is changing back or taking back control and giving it back to the rightholder where, for copyright law purposes, it is supposed to be.

But nevertheless I do think that you can argue and go a long way with the concerns that YouTube and platforms like them are currently raising. If you look at it from a free speech, innovation, user-generated content point of view, it is quite drastic.

KENNEALLY: And Jean-Baptiste, too, regardless who did the writing – Americans, Europeans, or lobbyists, whoever it was, you have a very interesting view regarding how well it was written. And it’s not very well written, I guess, from your perspective, but that doesn’t seem to bother you as much as I might expect it to. Tell us about that.

SOUFRON: I really don’t – we shouldn’t care. Just as was said by Tom, this is a try, and at some point, we need to try to find a solution to the mere conduit problem today because it doesn’t work anymore. We have so many troubles because of that, so we need to find solution. This is a proposal for a solution, and it’s not a bad one. The issue has been well written or not doesn’t really matter because at the end of the day I don’t think the text will be applied at all. As of today targets only two companies, Facebook and Twitter because YouTube already has Content ID and it’s super fine. So basically, well, you have Facebook and Twitter, they are the only two companies that will have to install a filter. The rest of the text is so protracted with arbitration clause, with a lot of little things that will – well, everybody will be interest to negotiation and to find a solution that will probably look like filters somehow. But they will have to negotiate and to find a solution. I don’t think the text will be applied as is. And actually, that’s a little bit what’s written in between the lines. Because at every corner at every step in the step in the text, well, you have (inaudible) to actually negotiate. So is it a text or is it like a treaty proposal somehow?

THIJM: I agree and also disagree a little bit. I do think that in practice that’s the intention of the provision that licensing agreements will be entered. But the danger here, of course, is like you have the patent trolls here in the US that you have copyright bounty hunters that ensure that their content is being uploaded for some purpose on the platform, and it’s immediately run to the bank. So for YouTube and Facebook and Twitter and also for Medium, Reddit, and other companies, SoundCloud, companies like that that is going to be a big issue.

KENNEALLY: We’ve been hearing throughout the day here at the Copyright Technology Conference about copyright law as an incentive to creators. And yet there is a sort of flip view of that, which is that is a disincentive to innovation. We live in an era when innovation is thoroughly prized. Talk about that and about the discord that seems to arise between copyright provisions for incenting creators, and then other types of copyright law which seems to be disincenting what is broadly called innovation. Do you have thoughts on that?

THIJM: Yeah, sure, I think you can look at innovation from two angles. You can look at innovation in the technology sector itself. As Tom already said that for YouTube, that’s not going to be a big problem because they have big pockets and they’ll probably enter into the licensing business for Content ID system. But it’s a problem for startups. It’s a big problem for startups. We know that the internet has a tendency of creating monopolies, especially in the tech area. Well, this legislation is going to protect them even more. So it’s not going to be helpful, I’m afraid, from a anti-competition point of view. Then, of course, there’s the innovation from the creator’s side. YouTube, in its campaign, refers to the song “Despacito” which was the song that was most played, most streams on YouTube ever on the platform. It gives it as an example because if you look at the “Despacito” video clip, they say we still haven’t been able to clear all the rights in that video clip. It’s absolutely impossible because there are too many rightholders. So they’re saying it’s going to be very difficult for the younger generation musicians, but also the YouTubers that are finding a way, making a living on YouTube to continue their business as long as they are using rights of others.

KENNEALLY: Jean-Baptiste, your thoughts on that discord between incentives and innovation as it applies to Article 13.

SOUFRON: I don’t know. I don’t think the text actually deals with that very well. Because at first maybe the text was too strict and would have actually been really slowing innovation or maybe even stopping it because it was too strict in the way it applied. Although today I think the limit for SMEs might be a little low. I was reading the text two or three days ago about what happens once you – the limit is at 10 million a year. So once you make more than 10 million a year, suddenly you’re not an SME anymore in the context of these texts. So what happens the first time? Well, all of a sudden, you’re 11 million. How much would it cost to then apply the text? And that’s probably one issue. We need – and what you said was very important because if we can interpret the text as, well, you will need to apply 1% of Content ID will be to be – you will be efficient enough. Then all of a sudden, what is the trouble? There is not trouble anymore. It’s something that you will have time to create your business and then to make it grow, and then slowly you will more and more cautious about copyright. Well, it’s not a bad way to think about it. I think in terms of incentive, at least, it will force people to try to think about the business model they will want to use. Plus if you look at many successful businesses today in the field of copyright, well, the one that – I’m thinking about the French competitor to Spotify, Deezer. Well, Deezer began as, yeah, a pirate platform. They began by putting online content without authorization, but they always tried to negotiate from the beginning.

KENNEALLY: And Tom Frederikse, last word on that topic. Just to put some numbers around this, I think that YouTube – Google, I should say, Alphabet, perhaps, to be technical – has published that they have spent something like $60 million to develop Content ID.

FREDERIKSE: Let’s remember what’s not changed, I think that’s important here. Copyright infringement is still copyright infringement. The author still owns his rights and he can sue for anybody that infringes it. YouTube has infringed copyright for 20 years. They don’t actually say that they don’t. They don’t defend and say they haven’t infringed, they just hid behind the immunity that’s in America called Safe Harbor, and in Europe and Britain known as the mere conduit exception, they’re very, very similar, and none of that changes. There’s no repeal to legislation going on here. So all that remains. This is simply a gloss over the legislation that suggests that there’s a slightly higher bar that needs to be met now, and we’re arguing about whether it should actually say that you have to use an upload filter or whether it should say funny things like use best efforts or measures or cooperation. I don’t think that really – if it’s not an upload filter and it’s only cooperation, I can’t believe that we’re going to have legislation that requires phone calls or chats or friendly exchanges. It’s just obviously not going to do anything. So it’s going to have to be an upload filter requirement, it’s going to have to be proportionate for large and small companies. That’s the only change. Everything else remains as it was, effectively.

KENNEALLY: Well, I want to thank the panel, Tom Frederikse, Chris Thijm, and Jean-Baptiste Soufron. Thank you all for being here. Thank you.