 Welcome to Building Tomorrow, a show that explores the ways tech innovation and entrepreneurship are creating a freer, wealthier, and more peaceful world. As always, I'm your host, Paul Matzko, and with me in the studio, Will Duffield, a research assistant here at Cato and editor of Libertarianism.org's prototype. This week, we're going to discuss a proposal from the European Union that would change how copyright law functions in relation to the internet. I promise you, it is way more interesting than that bare description sounds. To help us out in the conversation, we've asked Mike Maznick, founder of the tech news site TechDirt, to join us today. Welcome to the show, Mike. Hey, thanks for having me. Mike, you've written quite a few articles on TechDirt about the copyright directive, and we're going to get down in the weeds about some of the implications, some of the individual articles. But before we get there, why don't we start with the basics? So this legislation, it's proposed in the EU. Is the copyright directive, what's the point of the bill? What's it ostensibly supposed to do? So I mean, it's something that's been discussed for a really long time in the EU, and the original idea behind it actually makes some amount of sense. And it was a recognition that within all the different member states of the EU, there were all different copyright laws. And that actually many people believe, with probably good reason, was limiting the ability of various copyright related industries to develop all that broadly. Because if you could work on something that made sense under the copyright laws of France, but did not apply in Spain, that just became a problem. It had become difficult for, in particular, music and video services to develop across Europe or across the EU, specifically, because of the different laws in different places. So the original idea was behind the copyright directive, is can we set up a system that effectively would harmonize copyright law across the EU and hopefully then unleash new businesses and services based on copyright. And it makes sense. It'd be like if every state in the US had a different set of copyright rules, that would be a real mess. Right. Yeah. To some extent, there was a period when we had some of that. And I mean, it's interesting, this might be going a little off on a tangent very early in the podcast, but there is an argument that there is some value in actually doing it that way in that you do get experimentation. And you do begin to see like, if one state or country has very permissive copyright laws, does something more interesting develop there versus very lockdown copyright laws. But in practice, it hasn't really worked that way, especially just because of the nature of the internet and the fact that borders don't really matter that much on the internet, you do kind of want something that is more broadly applicable if you're going to build a kind of service or business that wants to be global. So yeah, I think the underlying concept of harmonizing the basic idea definitely does make sense. There's nothing wrong with it. And you see some of this tension in any move towards regulatory harmonization. The question then becomes, which way are you harmonizing? Right. Yeah. Exactly. That's a really good point where it's like, if you're harmonizing it in a way that is highly restrictive and problematic or that limits industry or that locks in certain winners and losers, that's a problem. If it really is opening up new opportunities for competition and innovation, then that is potentially more appealing. But which of those comes out of it is kind of the big question as always. And unfortunately, in lots of these processes, of course, certain industries are able to influence the way things go so that it does favor them and does hinder competition. So I think now we can probably break down some of the specific proposals. So as an entire policy package, there's lots of pieces that are just about the harmonization that are not overly stringent. But in looking at the criticisms of the kind of omnibus bill, there were two articles in particular that stood out for a lot of criticism. Articles 11 and 13. So why don't we go through those one at a time, what they propose, and why critics are worried about the implications of applying those articles to the entire EU. So article 11, that's the link tax. Do I have that right? Well, I mean, the people who wrote it and support it would take exception to that and say, it's not a link tax. Don't call it a link tax. But yes, critics of it are referring to it as a link tax. It could also be called a neighboring right or an ancillary right. That sounds a lot friendlier, doesn't it? Yeah, I mean, sure, a neighboring right certainly sounds friendly in some weird way. And how would that work functionally? Like, so this neighboring right, when it's applied, what would it actually do? Sure. So the idea is copyright is granted to whoever creates an original work, right? And then you can transfer that or assign that copyright to somebody else. When you are, if you're a publisher, say a newspaper publisher, you know, what normally happens is that, you know, you have reporters on staff and they are either on staff or freelance, but they are, generally speaking, assigning the copyright of the work to the publisher who holds the copyright on those works. And that is a system that, you know, has worked more or less for ever, you know, since for the last couple of centuries at least. But for some reason, a bunch of publishers have decided that in this modern internet era, that is no longer enough and they need an additional right and, you know, it is referred to as a neighboring right because it's not technically a copyright. It is just sort of neighboring copyright next door to copyright. And it is a separate right on top of the copyright, not for the creators of the article, not for the journalists, but rather for the publishers themselves over the contents. And the reasoning here or the thought process behind it is that there have been a number of news search engines or news aggregators with Google News being the most prominent and most well known that for some reason publishers, especially in the EU, and we don't seem to have this complaint in the US, but in the EU, a bunch of publishers have just gotten really upset about Google News and they feel that it is somehow unfair to them. And the unfairness, and I'm choosing my words carefully because I don't, I have a lot of difficulty seeing how Google News is unfair in any way, but I'm trying to be as fair as possible in my description of it. You know, Google News, you know, will aggregate a bunch of different stories together and link people to it. And with the news, it will show a headline and usually a snippet, which will be like, you know, the first two sentences or a couple sentences out of the first paragraph sometimes. And for some reason, the publishers feel that this is unfairly infringing on their rights, except there is no written right for that. You know, it's such a tiny amount of text that it's not violating the copyright. And it's also linking to them, thereby sending them traffic. But for whatever reason, which normally would be a good thing, right? Like you're driving, presumably most people read beyond the first two lines and want to arrive and end up arriving at the site of the copyright holder. And it's possible they'll click on the banner ads on the site of the publisher and etc. And we can actually see in the expressed preferences of the publishers that they actually know this and they recognize this and there are two ways in which we see that. One is that if they really, if they don't want to be in Google News, that's very easy. You can put up robots.txt and a little file on your website says to Google's scanning system, don't use these pages, right? So you can block Google News. It's very easy. It takes 10 seconds. So they could do that. They don't. The second, the second expressed preference that we see is that all of these publishers have search engine optimization people who work very hard to get their sites ranked higher in the search so that more people will find them through search and go to them and visit them in order to get traffic. So both of those things make it really clear that they actually do value the traffic that Google sends them. Yet, you know, what seems to be the case is they're sort of looking at and this is, you know, me who disagrees with their position, stepping in and stating what I believe their their real position is. So take that with whatever grant is all you want. They seem to be saying that like Google is making a lot of money. The publishers are not making as much money as they used to. Therefore, some of that money that Google is making should be ours. Therefore, we need this right whose only purpose is to say Google pay us in addition to sending us traffic. And so the nature of the right is that if you aggregate and post a snippet of text that you are linking to and you are doing that as a commercial entity that you have to license those works. In other words, you have to pay in order to link with a snippet. It's I mean, it seems like a case of legacy paper media versus upstart digital outlets and there's been a shift in both influence and in just finances. And it's because they're being left behind, there's a need to make up the difference in the funding flow. And I mean, I think we can sympathize with the plight of traditional legacy media, but the idea of taking out that frustration on these new aggregators is it's not channeling the energy in a very productive direction. And it's not going to help these legacy publishers in the end either, because if you make it more difficult for people to link to and therefore other users to access your work, fewer people see it, fewer people click on the ads and your revenues continue to decline. Now, Mike, one of your articles gave the examples of how this legislation was kind of tried out on the small scale. Again, in our kind of EU laboratory of states situation that in Germany and Spain, they rolled out similar legislation, but that the results were not what they the proponents might have expected. Can you explain that for our audience? Sure. Yeah. I mean, there's a long history with this and it actually starts before before the German experiment, which was a bunch of publishers in Belgium actually banded together to try and force Google to pay them. And that was a complete failure. I mean, Google sort of took them out of Google News and then they freaked out again, you know, express preference as opposed to, you know, stated preference. And, you know, and it just showed that, you know, how much they actually value the traffic from there. So in Germany, they passed this law that was similar to to what is in Article 11, saying that aggregators had to get a license in order to do this. And Google responded a couple other aggregators responded in different ways, but Google responded by leaving, leaving the German publishers in Google News, but removing any snippets with them, leaving snippets for other publications. So because the German text was a little bit different than the Article 11 text. And in the German text, it was very specific about, you know, headlines were OK, but snippets were not OK. The Article 11 text is not entirely clear on that. Article 11 text, it could be as little as a single word could be violating Article 11, which is why some people call it the link text. And that's because you see descriptions in the links. Exactly. Like if you have a word in the URL, that alone might be enough to trigger an Article 11 issue. So but but in Germany, at least they made it clear that the shortest of snippets wouldn't count, but but not not much longer than a very short snippet would count. So the way Google News responded was they they took out the snippet part. They left the headlines, but they took out the snippet of German publishers and they left in the snippets of non German publishers. And of course, that drove a lot more traffic to the non German publishers. And it significantly reduced the traffic to German publishers. And so the German publishers freaked out and they tried a whole bunch of things, including claiming that this was an antitrust violation, which was an interesting move, which failed pretty dismally in the courts. In fact, the courts came back and more or less or I don't remember. Sorry, I'm blanking. Now, if it was the courts or if it was like the the it's like a legislative committee or something. Yeah, like a German antitrust. Oh, agents, regulator, regulator who more or less came back and said, you know, if there's any antitrust issue here, it might be with you guys all colluding to try and force and sort of hinted at that and more or less said, like, if you continue to push this, we might investigate you rather than Google. And so they sort of backed off. And then what they did was they actually gave Google a free license. You know, the law required that you need a license. So Google ended up getting a license entirely for free. The snippets came back into Google News. And so now technically, I believe Google was the only one who actually got that license. So you have other aggregators who actually don't have a license. And so now what you've done is you've actually empowered Google to be an even, you know, more entrenched player. And so if your concern is antitrust, you've made the biggest, probably the biggest provider, more powerful. And the competitors are going to have to pay. I mean, it's going to be hard. And from the EU's perspective, a foreign search engine. Yeah, which is kind of incredible. And so then, you know, soon after that, Spain passed a similar law and Spain looked at what happened in Germany and said, well, we're going to correct the errors of the German one to make that not possible. And so they made it so that you couldn't get a free license. I forget the exact details of how they worded it, but they basically said, well, you know, it failed in Germany because of, you know, they were able to sort of force Google was able to force the hand and get this free license. And so you can't do that in Spain. So the Spanish law has some weird things, which also potentially, like, effectively makes like creative commons licenses, you know, illegal. It messes up all sorts of things, you know, basically, you know, assumes that everything has to be licensed for money. And Google respond to that by saying, well, we can't, we cannot operate Google news in Spain under this law and they shut down Google news in Spain. And so the result of that has been not great. And so there was a study that was done that looked at, you know, traffic to different news sources and in particular to smaller news sites and, you know, you know, upstart news sources. It's been tremendously damaging and they've gotten a lot less traffic because they no longer get any traffic from Google news. And so, you know, we have these two examples that show that this kind of law doesn't work or, you know, does more to entrench a Google and doesn't seem to to lead to any any money changing hands. Well, but you see like the the third or, well, fourth try. It must be the charm. Just keep staying in your head against the link tax wall and eventually go break through. I mean, there seems to be this this odd belief that like, well, if only the entire EU were covered, then magically it'll work. And yeah, that does not seem likely to be the case at all. Well, in looking to GDPR there, you covered all of the EU and you still saw niche services pop up specifically to block EU users from platforms hosted elsewhere who didn't want to render themselves compliant. Yeah, yeah, as well as, you know, to the benefit of larger established outfits who can afford the, you know, the cost of compliance. And there's kind of a corollary there to the Google news experience in Germany where the big players are better able to meet the regulations. So Article 11, the link tax. Article 13 then was this mandatory upload filter. What so far, listeners who don't when I say mandatory upload filter, it doesn't automatically evoke a concept in their mind. What are we talking about? How does an upload filter work? Yeah, so and just to clarify here, too, as well, because the supporters of Article 13 would argue that it's not a mandatory upload filter. And so another word there. There on Article 13. So Article 13 is basically saying that if you are, you know, commercial platform that has a lot of copyright covered content, which, you know, is basically anything thanks to the way copyright laws are today, almost anything as soon as it's created gets some level of copyright protection. But if you have a lot of it, you have to take appropriate measures to keep certain copyright covered content off of your platform. So if you're Instagram or Google or Facebook or whoever, you have to make sure that content produced by your users. So it's not your own. You're just the platform hosting it, but content uploaded by your users has to you have to take these appropriate measures to make sure it's not something that's copyrighted. Well, that's not unlicensed. Right. So because I mean, everything is going to be covered by copyright for the most part, you know, the vast, vast majority of stuff will have some sort of copyright, you know, even the photos that you take, you know, of, you know, your dog or whatever. You know, there's a copyright on that. And so, but, you know, so it requires these sort of appropriate measures. And that's sort of the language that they use to sort of try and claim that this does not require mandatory upload filters. But, you know, what everyone has said, and even sometimes the people who are supporting the bill will admit by accident is that they really mean they want everyone to have upload filters. And the idea here is that, you know, they've seen systems like mainly YouTube's content ID, which is this sort of filtering system that will look at a huge database that it has of works that copyright holders have given to YouTube and said, we don't want these things appearing on YouTube, or at least we want the ability to monetize them for ourselves or whatever and matches against that that database. And so they've the regulators have decided, well, this sort of kind of works for YouTube, therefore can work for absolutely everything. And therefore, every commercial service that hosts content for public consumption should be able to install some sort of filter that will make sure that copyright content that a copyright holder does not want on that platform can be blocked from appearing entirely. So now, you know, maybe for some of our listeners, they hear that they say, well, YouTube has this system, it works. If imperfectly, you know, what's the problem with, say, Disney putting up, you know, basically telling YouTube that they want any Disney movie or a clip longer than a certain period of a Disney film, it's going to be blocked, you know, there'll be a cease and desist letter sent. You know, it can be appealed, et cetera. But maybe our listeners might say, well, why is that such a bad idea? What are the unintended consequences of a content ID program? Even the best ones are the most effective ones. Yeah. So there are there are lots and lots of different consequences. And probably we could take more time than we have for this podcast discussing all the different ones. But, you know, you know, at a very basic level, first is just the the cost. You know, YouTube has has stated publicly that it costs them $60 million to build content ID. And that's just for for video. Right. So how many platforms out there are able to spend that kind of money? It's not that many. Right. I mean, there's four or five, maybe. And even then, I'm not even sure. And that's just for YouTube. That's just for for the video content and, you know, cover some musical content, so video and audio content. And that's also in a case where all those copyright holders have agreed to hand over, you know, examples of their their work. And so you have other platforms like Instagram that you mentioned before. How do you handle that question on Instagram? What if you go to a museum and take a photograph of a painting that is covered by copyright? Does Instagram have to be able to recognize that? How do you build a filter that recognizes that? Is it fair use? How do you build a filter that determines what is fair use? You know, the the the one that we have already in content ID is notoriously bad at figuring out what is fair use. You know, smaller platforms are going to have even more trouble with it. And then on top of that, you look at the number of stories out there of content ID completely messing up and pulling down perfectly legitimate stuff. It happens all the time. You know, we can certainly get reports and examples of it, you know, on an almost daily basis. And that's, you know, the most expensive, most sophisticated such system on probably the easiest content to make those calls on. And it and it still pulls down all sorts of of content and blocks it. I think my favorite example just out there, unintended consequence was there was a Family Guy episode in which they showed a clip from, you know, some like let's play YouTube user of an old like 1980s video game or something. And so the but then once Family Guy aired it, well, now the company that owns, you know, whatever the media conglomerate is behind Family Guy, it's now covered under their like their their copyright claim. They basically said, look, it was on the show. Therefore it's part of the show. Therefore you have to take down any use of it because it's, you know, it's it's taking off the show without license. And they, you know, there was under the US DMCA system. They they made the original person who they were actually borrowing it from take it off their YouTube channel, who had put up seven years prior to it showing up on the show. I mean, like there's this bizarre Urboros like devouring its own tail going on here, which is just, you know, and none of this was intended. In fact, no one even a lot of the people involved, including the creator of Family Guy was actually opposed to this. But once the machinery is in place, it's hard to resist even even when that's not part of your intent. Yeah. And I mean, part of it is just the fact that, you know, copyright itself, and this is something that I think a lot of people don't fully understand. Copyright is very, very context specific. Right. I mean, the exact same piece of work in two different contexts, one could be infringing and one could be not infringing. You know, it could be fair use, could be de minimis, could be parody, could be, you know, there's all sorts of reasons why the exact same use could be legit in one case and not the other. And that's something that, you know, as sophisticated as, you know, machine learning algorithms may be these days, they're not so good at figuring out the context of these things. And that gets also really problematic when you're dealing with the fact that, you know, these filtering decisions are going to be made, you know, pre upload, you know, if the idea being before you can even get them on the site, they, the filter is going to determine whether or not it can be published or not. And, you know, so we may not even know all sorts of stuff that is not allowed. And that's where the fear or the, you know, some people have been using the term censorship machines to describe the upload filters becomes a real problem. You know, if you don't even know if all these things are being blocked from being uploaded to the internet entirely, just because a machine has decided they must be infringing, even if they're not, what recourse is there? And it's not clear that there's any for some people. And that raises some serious concerns as well. So no filtering apparatus or machine learning system is going to keep up with the speed of culture, the extent to which or rate at which images, video, sound is remixed and recontextualized by masses of users spread out across the world. But it seems to me that underlying a lot of this debate is a lack of appreciation of scale. There's simply so much of this content being uploaded every day, every hour, that in order, even if you were to have a, you know, 99.7% effective filtering mechanism, which isn't something you're likely to approach anyway, you'd still then, given the millions of pieces of content that are uploaded, be taking down tens, if not hundreds of thousands of pieces of legitimate content simply in that 0.3%. Do you see the EU legislators as aware of that scale issue or is it something else? I think it's very difficult for most people to comprehend that kind of scale. I mean, I don't think our brains are designed or evolved to comprehend that kind of scale, right? I mean, we're used to dealing with a small group of people and small groups of things. The idea that, you know, I don't even know what the most accurate numbers are these days now, but like, you know, 100 hours of video are uploaded every second on YouTube or whatever, which is, I think that was the stat from like five years ago, so I'm sure it's much higher now. The idea that you can go through all of that in any reasonable way and not make mistakes, it's hard to comprehend. And I think it's, you know, it's a really, really difficult thing even for people who sort of live in this space and, you know, follow this stuff. I don't even know if I fully comprehend the size of the scale and, you know, and I've talked to people at these different companies and, you know, they just try and keep coming up with different ways to express it to you. And, you know, where it's just like, the idea that like, you know, you have to make, you know, on the order of, you know, five million decisions today, you know, five and a half million decisions tomorrow, six million decisions the next, you know, you are making so many decisions, so fast, it's impossible to do that correctly. You are going to make mistakes. And when you make just even a tiny few mistakes on such a large scale, then you're, you know, creating so much collateral damage. And we had a post recently, which, you know, you may have been referring to there where somebody had looked at, you know, the numbers if you were to get that kind of quality, you know, I think there were examples of like 99% accurate or 98.5% accurate. And then you sort of estimate like how much of the stuff that's being uploaded is infringing versus how much of it is not, you know, you end up taking down a ridiculous amount of perfectly legitimate content to really get at what is a relatively small problem. And that, you know, raises a whole other issue, which is, you know, probably a whole other podcast discussion not worth going into, which is like, you know, how much of this is really a problem? You have certain industries, you know, the music, the, you know, recording industries and the movie industries who continue to focus on this as a problem. And yet the evidence really suggests that this is not that big of a problem. Is there a lot of infringing content online? Absolutely. Are there a lot of people getting it for free? Absolutely. Is that a real problem for their business model? That's a much bigger, you know, bigger open question because, you know, what we're seeing is that when you have good, effective license services that people like to use that offer a good experience, the rate of piracy or the rate of relying on these services drops. And in fact, you know, the people who use these services over and over again, I mean, there's something like, you know, two dozen different studies showing this at this point, you know, those people tend to spend more money on the license services or on content or going out to see movies or on buying music or concerts or whatever, you know, the question is how much of a real problem is piracy on these platforms at this point? It's, you know, there's a lot of content there and that's what everyone focuses on. But is it an actual problem for the industry is a whole other question and will filters somehow solve that? Will filters suddenly make people buy more than they were before? That does not seem to be all that well supported by the evidence either. Well, we had that debate a couple of years ago here in the U.S. I was just thinking about it today in related, so it's the idea that piracy is a customer service problem. In this back during the debates over people torrenting large quantities of music and the pre-itunes, pre-easy internet access to television and video products, piracy was a huge problem. But then once, essentially once legacy TV film and music industries updated how they thought about the, how they made money off this process, the piracy problem became a tiny percentage of what it once was. So I was thinking about this today. I was one of the watch episode of the AMC show Preacher and they have the first episode of the season for free. So you get a taste of it and then sure enough went and bought the season, right? I'm someone who, and that's something that a lot of television film opposed doing at first, the idea of a snippet or a sample. They didn't wanna provide anything whatsoever for free and then there wasn't even a way to really legally buy it online. You had to go find, track down the DVD and wait a long period of time. But the model has shifted. So it feels a bit like there is the wave, the crest of the digital revolution is hitting these legacy institutions and they're desperately trying to paper over the cracks in the dyke rather than adapting their model to fit it. So it's interesting to see that happening over in the European Union right now. Now I saw from the coverage that this directive, the copyright directive was defeated in the EU parliament on a pretty close vote, 318 to 278 a few days ago. So Mike, should we pop a cork, sip some champagne, relax, are we all good now? No. So man, no happy ending. So the issue wasn't necessarily that it was rejected so much as what the EU parliament did was they agreed to actually open it up for possible amendments, right? So, and I don't, I'm in no way an expert on sort of EU regulatory process. So I've been learning some of this as we go as well but more or less, you know, the EU commission sort of created the directive, it got buy-in from, I forget what it's called but there's some sort of like committee of the different member states of the EU and they bought into it and then you had an EU parliament legislative affairs committee, which is, there's an acronym JURY which I forget what it stands for. They then voted for it and then normally what happens then is that sort of from there on the EU parliament would rubber stamp it. Instead, you had enough people in the EU parliament say like we should put this up to a full vote across the EU parliament as to whether or not it should be opened up for amendments. And so that was the big vote and that's the one that was narrowly approved which now effectively opens up the text so that anyone can amend it. Now, it was a really close vote and so there's all sorts of questions about what are those amendments going to look like they could fix the problems of Article 11 and Article 13 or they could make them worse or they could in the end remain as they were. And so there's gonna be basically a big fight over the next couple of months and then there'll be another vote in September on those amendments and then we'll see where these things go. And it seemed like there was a lot there was more support from I could tell among the MEPs for they were, fewer of them were as concerned about the link tax or about Article 11 and the neighboring right as we're concerned about the all necessary measures Article 13, like the way some of the votes broke down. So it's not hard to imagine given how close the vote was that in the amended version that wasn't quite as onerous that took out 13 or watered it down a bit would have a serious shot at passage. Yeah, and that doesn't even get a, I mean there are other articles in the directive that are potentially problematic and too in terms of like for search engines for like scraping data there's a text and data mining part of it that has some potential problems too but you know, there's so many things so many fires you can fight at once. A luxury of problems. That's right. But yeah, you're right that Article 13 is the one that, I mean, and to be clear like Article 13 is the one that's likely to have the most impact on general users of the internet that would have the most impact on kind of how the internet works and feels for everyday people. And so that would have the biggest impact and so we see that in terms of the protests and the focus of the protests and in the way that MEPs are seeing it. So Article 13 is definitely seen as the biggest problem and Article 11 has gotten significant attention but less attention than Article 13. So it is entirely possible that, you know, some sort of compromises worked out where Article 13 is removed or greatly watered down but Article 11 remains. And that to me would still be really problematic but you know, it's possible we could see all different kinds of permutations on that. So this European proposal would certainly affect US internet users or it to be put in place. Do we see anything like it making its way to the US or through the US legislative system? We seem to have been hacking away at intermediary liability with FOSTA not too long ago but are there any similar moves on the copyright front? Yes and no. So, you know, intermediate liability which, you know, for people who don't know is basically the question of, you know, if you're running a platform on the internet hence you are an intermediary. If somebody puts content that breaks some sort of law, who is responsible for it? Is it the person who created or placed that content or uploaded that content? Or is it the platform that is hosting that content? And, you know, generally speaking, we've always for I think very good reasons believe that it should be the liability and only should be on the person who uploaded or created that content rather than the platform that is hosting that content. It's, you know, perhaps a more extreme version of saying, you know, we don't blame AT&T when somebody makes a bomb threat using an AT&T phone, right? It seems to make sense to me. But, you know, that's been under attack globally for a very long time. And so the FOSTA debate was one example of that conveniently using a sort of hot button topic of sex trafficking in order to sort of drive a wedge into intermediary liability law in the U.S. But it seemed like if you're opposed this legislation you must approve of sex trafficking, right? It's the framing places. The accusations that have been made. Of course, now since that law has passed we've since seen and I've written a couple of articles now about how police departments, some of whom supported the passing of Cesta FOSTA are now reporting that it is harder, suddenly, magically, harder for them to find victims of sex trafficking. Because these websites are going offline and they used to use them to track down these people. And so, you know, suddenly there was just this report out of Indianapolis about how they're saying they just did their first arrest of someone in 2018 and they were saying, you know, last year was so much easier when we had back page to use to find all these people, it's like, you think, you know? That's what happens. And so, you know, but there's been this sort of coordinated attempt going back decades to sort of eat away at intermediary liability protections and indeed, you know, some of the same people who are working behind the scenes to get Cesta Pastor, the same people who are working behind the scenes on specifically on Article 13 in the EU. And so I think there's definitely a plan. And this is, you know, this also may be a potential tension but you can go back, you know, basically 25 years now and find examples of how particular industries, especially in the copyright realm, really, really like to use sort of this international strategy where you go to one region or country and you get them to pass some crazy idea to expand or extend copyright protections in one way or another. And then you work on some sort of, you know, international trade agreement to, again, harmonize between different countries, but, you know, it's always harmonized in one way or even better, you build this sort of ratchet system in where you get, you know, one country to pass something horrible and then you tell some other country region that they have to match and you get them to actually do it a little bit, even beyond that and you go back and forth and you keep getting people to push further and further. You get this sort of leapfrogging situation. And so there's definitely some indication of that. And there were a couple of sort of procedural moves with the EU Copyright Directive that certainly suggested or hinted at a plan to do something similar, get the EU Copyright Directive passed in there and then the different member states have to pass laws to apply it and then, you know, pressure the US to harmonize through some sort of trade agreement. And so, yeah, I mean, I think there's a real possibility that that is the plan. And that's sort of the guiding strategy behind all of this. Well, it's certainly something we need to keep on our radars as folks interested in tech, both here in America and we'll keep an eye on what happens in September with the EU Parliament. Mike, thanks for coming on and talking about this issue with us. I don't know how many folks who aren't regularly reading TechDirt would have picked up on how significant this issue was, which is a reminder to all of our listeners, check out TechDirt, find TechDirt's podcast in the iTunes search engine, give it a listen. Thank you guys for coming on and until next week, be well. Building Tomorrow is produced by Tess Terrible. If you enjoy our show, please rate, review and subscribe to us on iTunes or wherever you get your podcasts. To learn about Building Tomorrow or to discover other great podcasts, visit us on the web at libertarianism.org.