Tech Talk: Google v. Gonzalez & Twitter v. Tamneh — Talking Tech with Caitin Vogus & Jonathan Stray
Note: This podcast episode was recorded BEFORE the U.S. Supreme Court's recent decisions on the crucial online free expression cases: Gonzalez v. Google and Twitter v. Taamneh. Nevertheless, we've chosen to release this episode NOW because the conversation contains important information about the potential risks to users’ free speech that is more relevant than ever. <br><br>While the legal landscape may have shifted since this recording, the principles of protecting free expression and fostering an open and inclusive digital environment are not going away. In the Gonzalez and Taamneh opinions, the Supreme Court dodged any detailed consideration of Section 230 or the role of recommendation algorithms in enabling access to online speech, but we can expect that those topics will be back before the Court before long. So, let's dive into this discussion and keep advocating for the preservation of free speech in the online world.<br><br>The Center for Democracy & Technology – along with six other technologists with expertise in online recommendation systems – filed an amicus brief in the case of Gonzalez v. Google. The brief urges the U.S. Supreme Court to hold that Section 230’s liability shield applies to claims against interactive computer service providers based on their recommendation of third-party content, because those claims treat providers as publishers.<br><br>Here to talk a little bit about the amicus brief is Jonathan Stray, Senior Scientist at the Berkeley Center for Human-Compatible AI and Caitlin Vogus, (former) Deputy Director of CDT's Free Expression Project (currently Deputy Director of Advocacy at Freedom of the Press Foundation).<br><br>More on our host, Jamal: bit.ly/cdtjamal<br>More on Caitlin: cdt.org/staff/caitlin-vogus/<br>More on Jonathan: jonathanstray.com/me<br><br>(CDT relies on the generosity of donors like you. If you enjoyed this episode of Tech Talk, you can support it and our work at CDT by going to cdt.org/techtalk. Thank you for putting democracy and individual rights at the center of the digital revolution.)<br><br>Attribution: sounds used from Psykophobia, Taira Komori, BenKoning, Zabuhailo, bloomypetal, guitarguy1985, bmusic92, and offthesky of freesound.org.
Top Keywords
- section thirty 0.023
- section 0.015
- thirty 0.014
- court 0.013
- content 0.013
- case 0.011
- liability 0.009
- recommendation 0.008
- gonzales 0.008
- speech 0.008
- jonathan 0.008
- ruling 0.007
Transcript
Welcome to Tech Talk. Bye.
CT. Tea. Welcome to CDT's tech talk, where we dish on tech and Internet policy while also explaining what these policies mean to our daily lives. I'm Jamal Magdi, and it's time to talk tech. Before we get started, we want to let you know that the discussion you are about to hear was recorded before the US Supreme Court's recent decisions on the crucial online free expression cases, Gonzales v Google and Twitter v Tamna. Nevertheless, we've chosen to release this episode now because there's so much great information about the potential risk to users' free speech that is more relevant than ever. And while the legal landscape may have shifted since this recording, the principles of protecting free expression and fostering an open and inclusive digital environment are not going away. In the Gonzales and Tamna opinions, the Supreme Court dodged any detailed consideration of section two thirty or the role of recommendation algorithms in enabling access to online speech. But we can expect that those topics will be back before the court before long. So let's dive into this discussion and keep advocating for the preservation of free speech in the online world. The Center for Democracy and Technology, along with six other technologists with expertise in online recommendation systems, filed an amicus brief in the case of Gonzales versus Google. The brief urges the United States Supreme Court to hold that Section two thirty's Liability Shield applies to the claims against interactive computer service providers based on their recommendation of third party content, because those claims treat providers as publishers. Here to talk a little bit more about the Amicus Brief is Jonathan Stray, senior scientist at the Berkeley Center for Human Compatible AI, and Caitlin Vogus, deputy director of CBT's free expression project. Jonathan and Caitlin, thank you so much for joining us here today. Thanks for having us, Jamal. Thanks. It's a pleasure. Yeah. Happy to have you guys. So to kick us off, Caitlin, would you explain what the Gonzales case is all about? How was it related to another case the court has heard on this term, specifically Twitter versus Tamla?
Gonzales is a case about section two thirty and the recommendation of user generated content. So by way of background, section two thirty is a 1996 law that gives online service providers that host other people's speech a shield from liability for the speech of their users. So for example, section two thirty means that if a Facebook user says something defamatory about me, I can sue that user for defamation, but I can't sue Facebook. Facebook is shielded from liability by section two thirty. And in Gonzales, what the court is looking at specifically is whether section two thirty applies to content that a provider has recommended. The facts of the case are really tragic. The plaintiffs are family members of people who were killed in ISIS terrorist attacks, and they've sued a couple of online platforms. And what they're arguing in Gonzales is that YouTube violated a law called the antiterrorism act by recommending ISIS content to its users. Now they aren't claiming that the terrorists who killed their family members used YouTube or were actually inspired by YouTube content specifically, but more their argument is that YouTube is a tool that ISIS uses to recruit and spread its terrorist content and that it therefore should be liable under the anti terrorism act or ATA. And Google has responded, they own YouTube, and they've said, we are actually immune from this claim because of section two thirty. Sec, Twitter Vitamino, which you also asked about, is similar. In that case, family members are, suing Twitter, and they're arguing that Twitter violated the I ATA by, again, recommending ISIS content. But the difference is that because of the way the courts below decided the cases, section two thirty isn't the issue in Twitter v Tomna. Instead, in that case, what the court is considering is whether Twitter's recommendation of content is enough to make it potentially liable under the ATA itself. So in other words, assuming that section two thirty doesn't apply, what does a plaintiff need to claim in order to make out a case against an online provider for violating the ATA? So, basically, Gonzales is the case about recommendations in section two thirty, and Twitter v. Tomna is the case about recommendations and the anti terrorism act. And CDT filed amicus briefs in both cases because we wanted to make sure the court understood how its decisions could potentially impact online free expression and even depending on how the court rules could potentially limit free speech online.
And and in saying that, you mentioned that platforms fear of liability could limit free speech. Could you explain this a little more? Like, why would platforms fear, fear this or or or fear of liability?
So congress originally enacted section two thirty in part because it wanted to make sure that the Internet was a place where regular people could speak freely and find information. And the idea behind the law was that by limiting the potential liability for these online hosts of user speech, you reduce their incentives to remove user generated content. The reason is because if a host is worried that they might be legally liable for the things that their users say, they might say, I'm not gonna host user generated content at all. That's too risky. I'm not gonna do it. Or they might host user generated content, but really aggressively limit or moderate it. And we know that content moderation is imperfect. We know it often leads to the overremoval of totally innocuous or even beneficial speech. And even if platforms get it right 99% of the time, which they probably don't, because there are billions of pieces of content posted online every day, even that 1% error rate can lead to the removal of a lot of speech. And this can have disproportionate impacts on people from marginalized groups, especially, for example, l the LGBTQ community or women or people with disabilities and other speakers like that. And so we don't want a legal regime that incentivizes providers to even more aggressively moderate speech. And section two thirty, by removing the legal risk for providers, also removes that incentive to crack down on user speech. Gonzales specifically is about the speech that services recommend. So, again, it's important to think about what are the incentives that would be created if services were liable for any content they recommend. And because recommendations are so common and important on the Internet today, which I know is something Jonathan is gonna speak about, some providers might decide that the only choice they would have would be to, again, really aggressively moderate that content in order to reduce their risk of liability. And if you think about terrorist content specifically, providers might do things like remove news articles about terrorist attacks or even anti indoctrination materials because they mistakenly think it's terrorist content. So we're concerned about free expression because of the incentives that, the court's decision could create around over removals of content.
And so, Jonathan, turn to you now. What kind of recommendations are at issue in in Gonzales? And what role do recommendations play on the Internet today?
Right. Well okay. So this case, turns on the question of whether targeted recommendations are protected from liability under section two thirty. And it never really defines targeted recommendations. The the plaintiffs don't really explain what they think this is. But broadly speaking, many, many products and services have the problem of filtering down an enormous amount of available information, items they potentially could show to a user, to the small set that you actually see. So think about, for example, how many news articles do you actually read in a day? If you're a news junkie, that could be a dozen. But even a dozen is a tiny amount compared to the tens or hundreds of thousands of news articles published each day. So the idea there is that you want the computer to try to pick for you the things that might be most important or most valuable to you. And similar logic applies to your friends posting on social media, where you could have thousands of posts that you could see, but most of them you won't care about, jobs on LinkedIn, music on Spotify, movies on Netflix, and so on. So recommender systems are these systems used throughout the web to try to solve this problem of figuring out what you personally,
would would be most interested in seeing from the the huge pool of available items. So is it possible to distinguish between content that is recommended
online and or or content that's just being displayed? What's the difference, and how do we how do we separate those? Yeah. So this is one of the core questions in this court case. You know, if you read some of the other, briefs and arguments that have been presented to the court, there's there's sort of a little bit of a blurring between the different senses of recommendation. So, for example, if YouTube, wrote a review. Right? So YouTube, the company, or YouTube staff wrote a review of of a video and said, this video is fantastic. Everybody should watch it. I think it makes some really important points. I think everyone would agree that that was some sort of endorsement by YouTube. But on the other hand, what the plaintiffs are arguing in this case is that merely displaying the view the video, as an option for the user to click on, is a recommendation. And they make that argument saying, well, you're using information about the user and their past history and their interests and so forth to decide what they should show, that particular user. And they point out that these videos appear in labeled as for you. The problem is it becomes very hard to distinguish, you know, I make make some choice about what I think you see should see from, well, I'm just ordering things on a website in general. And there was a brief from Wikipedia that made this point. They said, well, you know, on the homepage of Wikipedia every day, we have, some highlighted articles, that, you know, the editors thought were particularly interesting or even a random pick, for example. And, you know, we can't be held liable for merely displaying content as an option for the user to click on. So the the case sort of ends up turning on, at what point is a platform sort of affirmatively endorsing the content. And we don't have very clear standards for that. Further, one of the distinctions that people try to draw is, you know, whether I asked for something specifically. So this is the sort of intuitive difference between web search, where you type in a query, and recommendation, where you don't. The the problem with making that the distinction is that, it sort of doesn't work in either either direction. So when you search for something like, say, I type politics podcast into spot Spotify, Spotify still has enormous latitude in inferring what it is that I mean by that search query. And so they have to somehow, filter and prioritize and rank the the items that might be politics podcasts, some of which might be material that, let's say, is, you know, damaging or harmful conspiracy theories. So it's not clear even that, search engines would be immune from, a law that induced liability for recommendations. On the other hand, it it sort of cuts the cuts too broadly in the other direction as well. If you have, a modern recommendation system accounts for a huge number of factors beyond the search query. Right? So your location and, you know, whether, it thinks you're interested in Python, the snake, or Python, the programming language. There's all of this other information that is used to decide what to show you. And it's not obvious why whether you typed a query at that moment, should be the thing that legal liability turns on.
Wow. So so, Caitlin, given what Jonathan explained about recommendations and the display of content, should the protection that section two thirty give providers for hosting user generated content change when they recommend it?
We think that section two thirty should apply to shield providers from liability for the content they recommend. In other words, we don't think it should change depending on whether or not providers recommend content. And part of the reason is because what the court is gonna look at in this case is the actual text of the law. Section two thirty c one is the provision at issue. And that part of the law says that it applies to any claim that seeks to treat a provider as a publisher. Claims that seek to treat a provider as a publisher are barred by section two thirty c one. And we think that recommendations of content are just another method of publication. They just reflect the provider's choices about how to order content, how to display content, and those are very similar to choices that traditional publishers have to make. If you think of, for example, a newspaper that's deciding whether to put a story on a one or a 26, the story that goes on a one is in a sense of recommendation that this is a really important story that newspaper readers should read. And we think it's similar when online providers are sorting content based on the various factors that Jonathan was describing. So we think that, section two thirty by its text itself does apply to the recommendation of content. But even more importantly, it's really important that it apply to protect free expression online. And I keep coming back to that and and and what we talked about earlier around the incentives of limiting or over removing content. And we really think that would be a huge problem if section two thirty didn't shield providers for any content that they recommend. And part of the reason for that is because recommendation is so, essential to what happens on the Internet today. It's really wrapped up in the display of content. And so providers would again have to resort to that aggressive limiting or moderation of content if they were afraid of potential liability, and that in turn would have really bad effects for user speech. They wouldn't just remove content that's posted by ISIS. They would also sweep up a lot of constitutionally protected speech too because they'd be worried that they might recommend it and it could lead to liability. So So those are the reasons that we think section two thirty should apply to providers even when they're recommending content.
Wow. Yeah. It sounds like there could be a lot of unintended consequences here if if if, not protected properly. Jonathan, you joined the advocates brief, CDT and other technologists filed that argues that section two thirty should protect the use of recommender systems. Why was this important for you to join this brief?
Well, I felt, and I think that some of the, other technologists who joined the brief, with us felt that it was important to have the voice of someone who has been involved in, developing and analyzing these systems for a long time, both in terms of just the background, how did we get here? What is the history of recommender systems? And that sort of thing. But also in terms of, people like me who study these systems spend a long time thinking about these questions of, you know, when are recommendations good or bad? And how should we, you know, design the recommenders? And how should we design policy and regulation to ensure that people are not suffering harmful outcomes? So I don't want to paint a picture that I don't think there are any problems with recommender systems. My research is all about the possibility that these systems may be amplifying divisive content and exacerbating, political polarization and other problems. So it's it's not that I think there isn't a problem here. It's that the way the court is talking about this case is sort of leading them in the direction of saying, well, the way we're going to decide which recommenders are safe and which aren't is on the basis of what kind of algorithm they use to show things to people. And I just don't think that's a very good way of of of making that distinction. In particular, I'm worried that, an adverse ruling in this case could make it harder for platforms to try new kinds of recommender systems that may have better outcomes or may do a better job addressing some of these types of harms. If If the liability ends up turning on whether the algorithm or the technique used to display content is sufficiently complicated in some way, so, you know, for example, most people imagine that reverse chronological listing, sort of classic Twitter mode, would not be considered a targeted recommendation in the way that the plaintiffs in this case are talking about it. Okay, so if that turns out to be the case, then that provides an incentive for everybody to use much, much simpler algorithms, which in some sense maybe is good because maybe those algorithms are easier to understand and we have a better sense of what's going on with them. But in some sense, maybe that's also very bad because it means that experimenting with new ways that might be better, it becomes legally risky.
And if I could just add to that quickly, it's really important, I think, to have amicus briefs from technical experts in a case like this because the supreme court justices are not technical experts themselves. And and they said as much at oral argument, justice Kagan said, we're not the nine greatest experts on the Internet. Should we be deciding this case? And one of the roles that an amicus brief like this one can play is bringing that expertise before the court. So the court has six of the greatest experts on the Internet, I think, in, CDT's amicus brief, teaching them about recommendation systems and in many of the other amicus briefs filed in the case. So that's a really important reason to bring that technical knowledge before the court to help, hopefully, educate it about the things that Jonathan was just talking about that can really matter when it's interpreting the law. Yeah. That's extremely important. And and since you mentioned it, and, Jonathan, I'll I'll start. I'll I'll come to you first, but, Caitlin, I would love to hear hear your thoughts, Zach, after. But we've all had a chance to hear the oral arguments in the case. Right?
What do what do the both of you think is missing, and where are the gaps? And lastly, what should the justices consider before making a ruling? Is is there anything that that we're leaving out?
Well, I was frustrated how little discussion there was over the question of what a targeted recommendation is. So this case is proposing to add a precedent that says, well, anything that is, quote, a targeted recommendation, might lose this liability shield. But what is that? And one of the things we tried to draw out in the brief was all of the things that targeted might mean. Right? There's all these different types of of information. So, for example, one of the things that's used to make recommendations is the item that you're currently looking at. Right? So you, you click a video on YouTube, and then based on that video, along the side, it says, well, here are some other videos that are kinda related to this one. So is that targeted? I mean, in some sense, it's using personal information that is, you know, what you're currently looking at. But I'm not sure if the if the justices would would intend that to be, considered targeted. Or another example, even sort of, you know, classic, let's call it Twitter classic. Right? The the the reverse chronological list of people that you followed. Well, you know, that's using a list of people that you selected. That is personalized to you. Is that targeted? And if that's not considered problematic, why is that? Is that because you had, you made some, explicit indication that you wanted to follow or subscribe to those people? Well, okay. But then the question of liability turns on, was the indication that I provided to the algorithm as to what I wanted to see explicit enough? And I mean, I think that's a legitimate question. Right? I think there's some great, arguments and discussions about whether you should base the things that you see on Facebook, for example, on whether you've clicked like on something or whether there should be more explicit controls about what you see. And that's a great conversation, but these are the sort of difficult questions that, really the justices didn't engage with at all in their oral arguments. And yet that is going to be the sort of, how many angels can dance on the head of the pin type question that people who build almost any type of app or product on the internet are going to have to be thinking about, if we get, a ruling here that says that targeted recommendations aren't covered.
Yeah. I agree that those sorts of definitional and line drawing questions, were not asked in as much depth as maybe would have been ideal in the oral argument. The justices were clearly trying to search for a line around recommendation systems and when they might be within or without section two thirty's protections, but, they didn't really come away, I think, with a clear sense of how to draw that line or what line is they would want to draw. I think, unfortunately, for me, there wasn't really much of a discussion of user speech rights in the oral arguments. The justices asked a lot of questions about how its ruling could impact the business interests of tech companies, and there seemed to be a an acknowledgment or an understanding that section two thirty is a fundamentally important law for the Internet itself. But none of the justices connected that back to the rights of everyday Internet users and how their speech could specifically be impacted by this case. And that's probably partially because Internet users weren't directly represented. You had the plaintiffs on the one side who were the family members of victims of terrorist attacks, and then you had the tech companies on the other side. And in the amicus briefs, you had some groups talking about user rights, like CDT's brief and and many other briefs. But the court just didn't pick up on that as much in the oral argument. So I hope that they will go back and look at those briefs when they're actually writing their opinion and and focus on that issue because I think it is the fundamental issue in the case, and the court really needs to take it seriously.
And I'll I'll also jump in here, and say that I'm I'm a little worried that people are fixating on social media when they discuss this case. And you see this in the oral arguments, and you see it also in a lot of the coverage. Right? So you get headlines and discussions that are, like, you know, deciding a a key case, you know, that will affect the future of social media, which is true. But a huge number of products that we use every day are also built on essentially the same type of technology. So for example, a news aggregator, if you go to Google News, that's a news recommender. What you are recommended on Netflix, the music that, you know, Spotify tries to to find for you that it thinks you might like, the jobs you see on LinkedIn, even web search. It's not clear that using the language that, the the plaintiffs and the and the justices have been using in this case, it's not clear that web search would be excluded at all. And the sort of worst case scenario is you get a ruling that makes search engines, think that they might be liable for how they order their search results. And so that would be very challenging as well to the idea that, well, some search algorithms would make you liable for recommending or sort of having in your search results, some piece of content that is, say, defamatory or, can be linked to terrorism or all of the, you know, terrible outcomes that can that can happen online. And that there's some the liability is going to turn on some difference about what type of algorithm you used, is very troubling.
So what should we expect to see from the courts in the coming month? Caitlin, would you take this first?
The court could issue its decision anytime between now and the June, which is the end of the term. And it didn't give a clear signal during oral argument about how it will rule, so it's hard to predict exactly what the decision could be. One thing I will definitely be looking out for is a decision where the court rules in favor of Google. It says section two thirty does provide immunity for the plaintiff's claim in this case, but it does so in a way that is still detrimental to online free speech in future cases. So kind of what Jonathan was saying, if it goes down the path of saying, well, some algorithms could make you liable and some wouldn't, and the liability really turns on some difference about what algorithm you're using. And in this case, Google's algorithm was was fine, but in a future case, a different type of algorithm wouldn't be. That could be really problematic in future cases. And so there's a scenario here where Google wins, but online free expression still loses. And I think we need to be on the lookout for that because it will matter for how section two thirty is interpreted down the road in the future. And it'll also really matter for congressional debates on how to amend section two thirty, which are very active right now. Congress is really thinking about whether it needs to change the law, and if so, how it should do it. And I think many in Congress are probably waiting to see exactly how the Supreme Court rules before they decide whether to take forward big changes or small tweaks to section two thirty or potentially leave it the same as it is?
So one way of reading, the oral arguments and the attitude of the justices is that they are searching for as narrow a ruling as possible. Pretty much every brief that was, submitted, on the, on the defendant's side, or I guess it's the respondents in the Supreme Court case. Anyway, on the on the, section two thirty should cover, should provide a liability shield for targeted recommendations side. Or even on the, the neutral side, you know, we're not taking a position, we're just trying to help you think about this side, pointed out that, a ruling here could have enormous consequences all across the Internet for a huge number of companies and and products and services. So I think they're trying to find a way to rule as narrowly as possible or to avoid ruling on it at all. And one way they could avoid ruling on it is by deciding in the other case, the Tamna case, that actually, you know, Twitter isn't liable for, terrorist content that appears on its platform under the ACA for whatever reason. And if that's the case, I think they could return to this case and say, well, actually, we don't have to decide it because no matter which way we decided, they wouldn't be liable anyway. So I think it's possible that we'll see them find some way to avoid making, a ruling here. Or if we do see a ruling, it'll be some incredibly narrow, legal point that they're ruling on.
Yeah. That's right, Jonathan. The court could avoid ruling on the section two thirty question if it decides in Twitter v. Tomna that regardless of whether section two thirty applies, the types of claims that the plaintiffs are making in these cases aren't enough to violate the Antiterrorism Act. So it's it would decide on the merits, basically, that the cases wouldn't go forward rather than on section two thirty grounds. I'm not confident that the court will do that because I think if you when you listen to the Twitter or vTOMNA oral argument, you could really hear the court struggling with the question in that case, which was around what kind of substantial assistance and knowledge does a tech company have to give to a terrorist group before it rises to the level of a potential violation of the ATA. And, also, I think that the court has signaled its interest in the section two thirty question for many years now. Several justices have written about how they want a chance at interpreting that law. And so I think it might be too tempting an opportunity for them to pass up. But we will have to see, I guess, what the court rules by the June.
Yeah. You know, I think we're in a sort of political environment where everybody wants to to take a piece out of the the social media platforms. Right? Everybody left and right is angry at them for something. Excuse me. So there's an appetite for doing something here. Unfortunately, I don't think this case is the, you know, the the the ideal vehicle for making good law, but they may feel pressure to, rule on it anyway.
And before I I this has been a great conversation. So so thank you both. And to just close this out, I I would love to hear Jonathan starting with you and then Caitlin, moving on to you. Any final thoughts, on on this case or or Twitter v. Tanya?
Well, I'll try to take a somewhat broader perspective here and say that while I don't think, a reinterpretation of this law is a good way to try to address some of the harms that, large online platforms can have. I think it's a real issue that we do have to address. And so I think really what it's going to turn on is sort of the public debate that we've already been having for several years now about what exactly is their responsibility towards individual people and towards society as a whole, and what kind of standards should we hold these systems to? And And I think that's really where the debate is stalled right now. Everybody agrees that the status quo is kind of untenable, and there should be some sort of accountability for a wide range of adverse outcomes. However, there's much less agreement on, for example, exactly how we measure those outcomes. Right? How if we're worried about, let's say, social media and mental health, can we, first of all, get a measure of how bad the problem is? Could we set standards for, sort of the, magnitude of the bad outcomes or or, probability of the bad outcomes? Because the reality is that these are very large systems used by, in some cases, billions of people. As one platform, engineer put it to me, you know, when you have 2,000,000,000 users, anything that can happen does happen. So much like, setting standards for pollution or fighting crime, you never expect to get to zero. I don't think it's possible to build a system like this where nothing bad ever happens. So then we have to have a conversation about, well, how do we think about the trade offs between the value that these things produce for many people and the harms that they produce for others? How do we quantify that? How do we set standards about that? How do we think about accountability? And that's an extremely important discussion. I just don't think that this court case is the right place to be having that discussion. But, for example, I would love to see congress pick it up.
I agree, Jonathan, that people are dissatisfied with the status quo, and that is a huge part of what's going on in this case and also motivating congressional conversations around section two thirty. And congress has been debating section two thirty for years now and has not been able to come up with a bipartisan consensus on how to change it. And we've now seen the court also searching, I think, in vain for potential reinterpretations of section two thirty. I think in terms of potential solutions, one thing our amicus brief talks about are the other avenues for addressing some, not all, but some concerns around social media in ways that wouldn't impact user speech. And these are things like enacting comprehensive federal privacy legislation that would limit the types and amount of data on individuals that services can collect and then use for these recommendation algorithms and other purposes in ways that I think some people find extremely troubling. Or another avenue potentially could be around competition law and ensuring that new services can enter the market and effectively compete with the incumbent big tech companies. So there's more user choice around the types of services they want to use, and hopefully, we have a healthier and better online information environment as a result. With respect to section two thirty, only time will tell about whether the court or congress is going to make changes to it, but I do think the important thing that they need to understand, either judges or policymakers, is that any changes they make around that law specifically could really impact free speech. And the Internet is such a powerful tool for regular people to use in order to speak and to find information, and we need the law to make sure that it stays that way. So I hope that congress and the court are keeping that in mind when they're thinking about section two thirty.
Well, Jonathan and Caitlin, thank you so much for joining us here today. It's been a pleasure having you. Thank you. Thanks, Jamal. And for all of our listeners, to keep up with the work of CDT's policy teams, please visit us at cdt.org or follow us on Twitter, Facebook, Mastodon, and LinkedIn at SymDemTech. This is Jamal Magby, and thank you for talking tech.
Hi. I'm Riddhi Shetty. I work on the privacy and data project here at CDT. Recently, we've been advocating for stronger federal and state guidance and regulations against consumer data harms that limit economic opportunity. You can support this and all we do here at CDT by going to c d t dot org slash techtalk and donating. Every donation matters. Thank you for enhancing civil rights and civil liberties in the digital age.
