Defamation Law in the Age of AI with Lyrissa Lidsky

- Welcome to Free Speech Unmuted. I I am one of your co-hosts, Eugene Volokh at the Hoover Institution at Stanford University. I'm also professor of Law Emeritus at UCLA Law School. My other co-host is Jane Bambauer, who's a professor at University of Florida Law School, and we are delighted to have on, on this episode, Jane's colleague Lyrissa Lidsky, who's also a professor at University of Florida Law School. But in addition to that, she is one of the reporters for the Restatement third of Torts, defamation and Privacy. That turns out to be a big deal, but, but some of you may be wondering what is the Restatement, what does it mean Restatement, and what are reporters for it? I thought there was a, that she was a professor, not a reporter. So we're gonna figure all that out and then we're gonna see what, what she has to tell us. It turns out to be a lot about the Law of Defamation. So, Lyrissa, tell us, what is the Restatement of Torts and what does it mean to be a reporter on it?

- So, the Restatement of Torts is put out by something called the American Law Institute, which is a body of esteem practitioner, legal practitioners, professors, judges. You have to be elected to be a member of the American Law Institute. And then periodically, really, once a generation, they decide to engage in a process where they look at the law in all 50 states and they decide to restate that law, which, which basically means making a clear statement of what, usually it's what most states do, the most articulate, well reasoned statement of what most states do, and if, if necessary, to, to synthesize and to modernize and to make a clear statement involving US tort law, aspects of tort law. And it's heavily relied on the document. So, so who uses a document like that? So the document is used by courts when they have a gap in their law and they wanna look to best principles to fill it. It's particularly used by federal judges who decide a lot of cases and look for an authoritative source of interpreting state law. It's used by students who wanna understand a whole body of law. It, it contains rationales for all the rules, but the restatement really is, so the, the last restatement of defamation was done in 1977, and as you might know, 1977 was really before the internet revolution and before social media and before generative artificial intelligence. And so the law of defamation, which is about protection of reputation and balancing free speech concerns needed to be modernized to reflect the internet revolution. And states have been working out what to do about that. And so the reporters go in and they look across all 50 states and see, you know, what the, what the law is and what the best account of the law is,

- Right? So it's a tremendously important resource as, as you point out for lawyers, for judges, for law students. When I, when I used to teach torts, I would often teach using sections from the restatement. They often come with very helpful. In fact, they always come with helpful comments and illustrations. Obviously, the American Law Institute is a shadowy organization conspiring to No, no, no, no. It is a, it is an organization that is not a governmental organization. It can't make something law, but judges often find it highly influential. A restatement of torts has been extremely influential, but there have been other restatements of contracts, of, of conflicts, of laws, of various other subjects. And, and, and judges will often say, and judges do in our system, make law oversimplifies things, but basically make law, they often say, look, you know, when there's some uncertainty about what the law ought to be in our state, we're going to adopt the well-reasoned perspective of the, of the restatement. And although of course, they're free to say, Nope, we just don't agree with it. We like to do things our own way. And there are many, many people involved in the process. There's a members consultative group on the, on each of the restatements and such. But the reporters are the people, I think almost invariably academics who are given the charge of taking the lead in drafting subject to feedback, which they take very seriously from other members of the institute. So in any event, tremendously interesting and extremely time consuming process. It literally takes years to prepare, to prepare arrested many years. But, but very important. And I think it's, my sense is it's the way law ought to operate. Obviously the law is made by people who have the legal authority to do that, either legislators or judges, or sometimes voters voting on initiatives or constitutional provisions or whatever else. But the proposals are often drafted by people who have actually studied the subject for their whole careers. So, so Larissa, tell us what are the, the kind of the most interesting issues that you've run into in the restatement process? The ones where the previous restatement from back in the 1970s, what was cutting edge at the time just hasn't kept up. And that there are things we need to think about either because of technological change or just change the way the, the, that the legal system has been operating.

- So obviously, as you said, 1977 is a long time ago, and oh, one of the biggest issues ha has been

- Jane, were, were you, were you alive in 1977?

- Thank you. I was, I was, I wasn't very old, but I was,

- I I was, I was, I was the, the advanced stage of nine at the

- Time, I was also nine. So

- There you go. Jane, I think is the baby in this group.

- Yes, yes, I was, I was just a, a, you know, a, a a a a wink between a, a man and a woman. Not yet. Anyway. Okay, there you go.

- So, so one thing, one thing I would want people to know is that it's not like a couple of law. I have a co reporter, Robert Post, former deed of Yale, and now professor at Yale. And I, I wouldn't want your listeners to think that just the two of us concocted ourselves in this, whatever we say goes, because in the, in the process of making this, you know, best statement of law is we do have a board of advisors who are all specialists in the field, and they come from different parts of, of the legal profession. Some of them usually defend p or we usually represent plaintiffs. And some, some of them usually represent defendants, and some of them are judges, some of them are professors. And so they advise us, and as you say, we take a lot of their comments very seriously. And then there's like a, a council body that has to approve the draft after that. And then if they approve the draft, it goes to the 2000 plus body of the membership for a vote. And so it really is a, a big responsibility to think about coming up with a best account that also is going to, you know, go through this, this rigorous, rigorous vetting pro process before it gets finalized. So you were asking what issues?

- Yeah,

- Yeah.

- We've done the procedure, of course, we know the lawyers through superpowers to turn every question to the question about procedure. And, and obviously procedure is important, and here the, the kind of the law making not quite procedure here, but what about the substance? Like what, what are the, the big substantive issues that you folks have had to deal with?

- So there's really two big substantive issues. Okay. That maybe three if you, if you count one of them that's still in the works. So one of the most significant issues in defamation law that has been talked about, you know, going back for hundreds of years, is it, it's a very complex body of law. And it's complex in part because of its history. Its started life as a sin. It, it could be judged in the, you know, slander could be judged in the ecclesiastical or church courts. And so you could have to do penance for slandering somebody. And then it, it became, you know, kind of a civil action that you brought, you know, you could instead of go having to go out and fight a dual or, or kill somebody who damaged your reputation, you could get money damages for that. And then at various points in its history, it's been a crime. It used to be a crime to speak ill of the king or the great nobles of the realm, the piers of the realm. And so all of those different periods in defamation history have left complexities in the bottom body of law. But one of them is the distinction between slander, which is usually defined as oral defamation, you know, words spoken versus libel, which is written defamation. And there's some complicated doctrines that go along with whether a statement is slander and whether it's libel and it affects what damages are recoverable. It depends on what topic the slander's on. And so for hundreds of years, commentators have lamented some of the artificial divisions that flow out of the distinction between libel and slander. And the last generation, when this, when this project was done in 1977, the reporters reluctantly kept the distinction between libel and slander, even though it had already become artificial as a result of technology. You know, for example, is a live broadcast slander or given its permanence that it's in a fixed medium, potentially? Is it libel? And those kind technology had already made those issues difficult to resolve, but the internet has just made them, you know, it has made it look obsolete to draw a distinction between written defamatory statements and oral defamatory statements and all the various little nuances of doctrine that come along with that. So one of the things we're hoping that the membership will approve in this draft of the restatement is to do away with the distinction between libel and slander and all the little, you know, the specialty terms like libel per se, and libel per quad and slander per se, and slander per quad and special damages. If the slanderous statement only falls within one of four categories and, and all of the attendant complexity. Nice.

- Normally I'd ask you to help the audience understand what all those categories mean. But I'm actually not going to do that in the hope the rest statement will become supreme.

- Yeah, yeah, I would, I mean, I can explain all of them because of course you become a specialist in this field. Part of what makes you a specialist is you can go down every rabbit trail into every nuance. But some of these nuances really are just artifacts of history that have long outlived any rationale they had, for example, the distinction, there's something called slander per se, and if slander it falls into certain categories, you don't have to prove special damages, you can just get damages more easily. But those, those categories come from a jurisdictional battle between the church courts and the, the courts of the, the peers, right. The al courts. And so the one, so the, the four categories were thought to be the kinds of things that are most likely to cause money damages and therefore give rise to a civil action as opposed to things that maybe aren't as likely to cause money, data, images. And so they can just remain a sin. So, but, but

- You know, like, so well, and then also this distinction between liable and slander. I mean, you were saying that the internet makes us question why we're asking it, like to just, just to, you know, to give an example, I, I like teaching this case where one model made two, two models supermodels or whatever, models not supermodels, were in a dispute. And one made a website that apparently nobody except the other model saw, and it was just full of nasty posts about the other model. And one of them said she was a whore or something like that. And so, you know, and so, so there, the, the, because of the nature of, you know, the, the court took for granted that it was liable, but was wondering Yeah. But was wondering whether a whore is a statement of fact and stuff, but like the fact that we treat that as liable as one of those more serious cases, even though the internet, most of the internet isn't really visited more than a couple times, you know, I think goes to show that the assumption that written, you know, written words have a longer life and a longer reach just isn't true anymore.

- Yeah. I mean, it's an artifact that doctrine is an artifact of the printing press. You know, that was the new technology when Oh yeah, we take things that are in writing much more seriously. I don't think we do. I really don't think we do. And so they need to be harmonized and unified. And, and you're, you're right, when the internet came along, most courts said, you know, this old slander stuff is too arcane. We're just gonna call it libel and, and call it a day without a whole lot of analysis. There was some, you know, professors at the time saying, you should really treat it like slander. 'cause people don't think about it much before they say it. And

- Right, yeah,

- Yeah. I that libel.

- So - Snapchat is the ultimate test of, of these things. Like, is a Snapchat a slander or libel? They just call it libel,

- Right? And just, just for further evidence that this is still a live issue, it's still causing, literally yesterday as we record this, and by the way, I mean literally in the sense of literally, and not literally in the sense of figuratively. So literally yesterday as we record this, I was reading a case called Walker v Cipher from Eastern District of Tennessee Federal District court in Tennessee, where the question was, here's what the court says, quote, the issue before the court, therefore is whether an alleged oral defamatory statement when later published online as a video, constitutes a plausible claim for liable under Tennessee law. Both parties agree that no Tennessee Court has directly addressed this issue. So this is still Bedeviling courts and, you know, they maybe reach reasonable resolutions, but it takes time and effort and money of on the part of the clients who have to pay the lawyers. So, so yeah, it makes a lot of sense that this could be something that's simplified. So that's one issue. Larissa, what other, what other kind of big picture questions are you folks dealing with in, in the rest statement and maybe some big picture ones you're not dealing

- With? Okay, so, so one of the biggest picture issues we're not dealing with is, so the restatements have historically been about tort law, okay? And people may or may not understand that defamation is both a tort, meaning you can have a civil action for monetary damages for wrongful conduct that harms your reputation. But it is all a tort that is underpinned with constitutional law with the First Amendment. Okay? And so there's, so states, states are the ones who give you remedies for, harms your reputation, but they have to frame their tort law with sensitivity to free speech concerns. And so their rules have to adhere to constitutional rules. So the restatement is only concerned with the state tort law, and it can't really cha it couldn't change if it wanted to constitutional law. That's not its mandate. And also that's the, you know, the Supreme Court are the ones who decide that. And so we have to accept the current constitutional principles as a backdrop against which to draft these, these black letter tort rules and explanations and justifications for the black letter tort rules. The other thing that we anticipate, so tort law not only has a constitutional aspect, it also has been modified at times by statutes. So one example is a lot of states have what are called anti SLAPP statutes designed to give quick procedural rep resolution of defamation cases so that, you know, it makes it harder to use bringing a defamation suit to punish the, and chill the speech of your critics. Okay? But the biggest kind of statutory change that reflects the internet is section two 30 of the Communications Decency Act provides immunity for internet service providers, big tech platforms for user-generated defamation on their platforms or on their, on their services. And that's been huge, you know, it was a law passed in the nineties to, to fuel the growth of the internet and fuel the growth of, of the platforms. And some people are dissatisfied with it because they say it means that platforms don't have to police content, they don't have to be responsible in rooting out things like hate speech, defamatory speech, harmful speech. And so this, this one question would be, what is the proper rule if the big tech platforms or the internet service providers no longer have that immunity? Okay, so the common law kind of stopped developing in the 19 nine, in the late 1990s when that statute came into play, because we didn't have to place big tech platforms any longer into preexisting categories and asked, are they like a newspaper? So as a, are they a publisher? Or instead, are they more like a a a telephone or are they more like a telegraph? Or, you know, what, are they more like a, a newspaper vendor? What are they exactly in terms of their responsibility for third parties who use them to commit defamation?

- Now, some of that question in a way is probably going to be pressed sooner rather than later because of the defamation claims brought against LLM chatbot chatbot companies, right? Like these, I know both of you have written about libel in the context of large language model output. So I imagine that's, is that's something that the a LI did not touch yet. Right? Okay.

- No, so, so the a LI like laid out principles, but the actual application, we try not to get too far ahead of anything that's been cited in existing law. And so given how new this is, we lay out the principles that could be used to resolve it, but we don't have a 50 state survey of how people have treated the

- Yeah.

- You know, chat bots yet, so,

- Well, I don't think we could have one yet, but

- Yeah, we have one decision, but this is

- One. Yeah. Yeah. But this is a, this is an area where section two 30 would seem not to apply. I think all of us in our writing have made that assumption, because it is not a platform that merely hosts other human users gener generated content. And so, and so I think we'll see courts have to grapple with exactly this, this, well, this among other questions of how to treat a new technology company when they're kind of operating at a scale that even a distributor and a publisher didn't have to work at. Right?

- Yeah. Yeah. So, so one of the things about the immunity under the, the Communications Decency Act statute is it kind of had a notion that conduits should be, that they're acting like conduits, they should be protected. But the statute has specific language that I don't think applies, as you say to generative ai because they're not just carrying content, they're creating content in a sense. So they're not

- They are generating content.

- They're generating content. Exactly, exactly. So I don't think they're con content, if

- The lawyers have been involved earlier, they probably would've not used the term generative ai. Oh,

- Yeah, exactly, exactly. They should have had the lawyer. Yeah, exactly. I don't think they're conduits in the, in the terms of the statute, but I do think they have conduit like aspects that should be taken account in deciding what the tort rules are governing them. And we've got a policy choice to make in choosing which tort category to apply to them. Most legal reasoning, especially in torts, is reasoning by analogy. And so the question is, what is the proper analogy? Are they, again, are they like the newspaper publisher or they like the news magazine distributor? And I, I, I kind of think they're quasi distributor, like, but of course in, in kind of pushing them toward that category, I'm reflecting, you know, my own thoughts about what the proper policy is, the policy behind the category.

- Well, and even distributors maybe are at, you know, maybe even that analogy starts to fall apart, because as you, as you and Drew Daves wrote, wrote recently or soon to be published in Journal of Free Speech Law, a distributor once told that something is false can remove it. The removal or the, you know, ma ma making, making false statements about individuals, making them, you know, making that impo, you know, sort of making the guardrails so that that does not happen seems to be quite complex task. Okay. Eugene, I know you've written about this too, so please cut in.

- Well, I wanna move a little bit to something that is, as Larissa said, is outside of her remit as, as a report for the Restatement. And that is this big picture First Amendment question that, that people have been a buzz about for, well, in a sense for more than 60 years, but especially in recent years. This question, which is sometimes shorthanded, is, will New York Times v Sullivan be overruled and should it be overruled? But I wanna break down into three sub-questions, and there are many more sub-questions we could ask. But the Supreme Court has basically made it a lot harder for plaintiffs to win defamation cases by requiring that they show not just careless falsehood, but knowing falsehood or reckless falsehood. Basically deliberate lie in at least three kinds of cases. One is public officials, people who are public officials who are, excuse me, people who are public officials who are suing their critics or closely related area candidates for public office who are suing their critics. That was New York Times v Sullivan itself, because Sullivan was a, was a, a a, a local official, a second category, so-called public figures. They could be athletes, they could be entertainers, they could be business people, although they have to be, generally speaking, really very famous. And these days they're really important people, maybe with some exceptions. Elon Musk is super famous. Jeff Bezos is super famous, but a lot of very important business people are not super famous. Whereas a lot of important, a a lot of a, excuse me, in my view, less important athletes and entertainers are famous. So they too, if they are to win a defamation case, have to show, have to show deliberate lie and not just careless falsehood. And there's a third category, which is limited purpose public figures. These are often people who are not famous at all, but who get involved in some public debate. Let's say they get involved in some criticism of a government official or some challenge to some proposed policy or some ballot challenge to some ballot measure or some such. And then they become public figures for purposes of, of allegations about them that in some measure are relevant to this public debate. And that, in a sense, has always struck me as a, as a particularly, particularly troublesome broadening of the public figure category, in part because if you were to say, if you get involved in a public debate, you have to pay a tax, it's the public debate tax, you have to pay a hundred dollars in tax. We'd say, well, that's clearly unconstitutional. That's a content based penalty on your engaging in First Amendment rights. Excuse me, your exercising your First Amendment rights. But if we say, well, as a condition of getting involved in public debate, you surrender a considerable part of your right to sue for defamation. That's not only not unconstitutional. Supreme Court tells us that's constitutionally mandated, even though that too is in a sense a content-based tax. Like many people would say, I'd rather pay a hundred bucks rather than have it be open season, more or less on people, people saying false things about me, the damage, my reputation. So those are three, at least conceptually different categories. So one question is, do we think the Supreme Court within the coming decade, let's say, is going to revisit the supreme, the its earlier conclusions that in all of these areas the plaintiff loses unless the plaintiff can show deliberate lie knowing or reckless falsehood. And the second question is, do we think the court should revisit that? Maybe we should have the law be more defamation friendly and at least some of these categories? So I wanted to hear what Larissa had to say, Jane, I'd love to hear what you have to say as well about it to foreshadow, by the way, as to what I have to say. I don't know. I think these are really hard questions.

- Well, so let me say, I'm glad, I'm glad you opened it up, because as you say, New York Times calling for reversing New York Times versus Sullivan is really a shorthand. And I think a lot of the people that call for that don't specify what it is that they want, because New York Times versus Sullivan was the first step on a long path towards systematic constitutional of the tort of defamation. And so New York Times versus Sullivan was the easiest step on the path because it dealt with a public official, it dealt with weaponization of defamation law to try to silence journalism on the most important issue of the day, which was the Civil rights movement. And the, the case that this, the, the Supreme Court decided there were minor technical errors in the article that the New York Times published, but it was really had the, had the, had the case come out differently. It could be that we wouldn't have a New York Times today. Okay? So the, that the, the southern officials were using this to quell outside newspapers like the New York Times and the LA Times from reporting on the Civil Rights movement. And it would've killed that reporting fundamentally, and it could have sent them into bankruptcy at the time, given the magnitude of the damages that were a stake. Okay? So when people call for the end of New York Times versus Sullivan, often, I think they're really thinking about this public figure extension. So it's one thing to say, if you enter the public arena as a public official, you gotta have thick skin. And in country where the people are really the source of power, where we have popular sovereignty and democratic self-governance, the people need to be able to critique those who rule them in very heated ways. And they need protection against being bullied out of critiquing their public officials in those ways. And so the Supreme Court, the logic of that says, well, there's a lot of really, really powerful people in society who aren't public officials. Think of Elon Musk, for example, who's more powerful, you know, some ma the mayor of Gainesville, Florida, or Elon Musk. I think that's an easy answer. No, no, no diss on Harvey Ward. But anyway, I think that's an easy answer. And so the Supreme Court said, okay, these, these people that are household names, I always say you can tell a, a real public figure if they, if they have a one word name that's recognizable like Cher or Madonna. Madonna, you can tell I'm old by the examples I use Brittany Taylor.

- I'm sorry. You you mean, you mean the Blessed Virgin Mary?

- Yeah, exactly. That those are always public figures, but it was this extension to other people that I think is the most controversial aspect of New York Times versus Sullivan for a lot of people. And what I really think is it would not be the end of the world if they reigned it in a little bit. And in fact, I think if you just look at Supreme Court decisions defining the public figure category in cases like timing versus Firestone, it's very clear that the Supreme Court meant the, meant the category to be narrower than it has been interpreted by lower courts. So, so I, I think if they're gonna do anything, they would go back and say, no, we really meant it. The public figure category should be really narrow.

- Yeah. Okay.

- Can I, can I then, yeah.

- Well, so given that Eugene made this provocative chilling effects kind of argument, and that Larissa, you are at least tentatively okay with reigning in the category within right now, you don't, you don't love it, but you're okay with, right?

- Yeah. - Well, so, so let me just lay out the argument that this would be very, very bad, especially for speakers. I think that you'll, you know, I, I get the point Eugene made that you run the risk that if you become a public figure by speaking out about something, you are going to have a target on your back. On the other hand, who in their right mind would say something about somebody else in order to make a public statement, in order to engage in the controversy, if they knew that a jury could use, not just, you know, could, could, could, could, could, you know, if, if they happen to be wrong, if they happen to be saying something false and a jury is asked only whether they should have known rather than whether they did in fact know that it was false. But, you know, I think people would feel pretty anxious about, about engaging. And also maybe they'd also feel more anxious about things that might ordinarily go into other categories, like whether, you know, what they're saying is really of an opinion or a statement, all of that. A lot of that we don't have to get to those hard questions if we have the malice standard to help us just sort of wash those cases out.

- One thing that's worth noting, so one of the things I do when I'm wearing my defamation hat is I write a treatise or I, I become the author of a treatise that it has to be updated every year with every defamation case in, in the appellate courts in the country. It's called, it used to be called Sac on Defamation. It's gonna be linsky on defamation.

- Woo. Oh, that's so cool.

- Oh, yeah. Still didn't know that. But anyway, I have it on

- My desk, so, you know, I can't

- Wait to do it needs to be rewritten, but I've been doing the update for the past two years, and so I read every appellate case on defamation every year and have to summarize any of them that, that break new ground. But one of the things I've learned from doing this is how many cases are still one purely on tort privilege grounds?

- Oh wow. Yeah.

- And one of the things I think is probably true about the Supreme Court's systematic entry into the field of defamation is that it has blocked some developments in state courts in tort law that might actually, you know, get to the same ends. This kind of followed on from Jane's comment that, that there are, there are various ways to win a defamation case. And if you close one door, another door opens. And so these, these tort privileges remain a very, very robust means to protect defendants.

- Right. And just to lay things out to, to, to, to the listeners, 'cause it's easy for this kind of stuff. Among, among three people who know live a law to, to, to turn into, into a little bit of, of kind of inside baseball is that defamation law has long provided so-called privileges, which protect speakers in certain kinds of situations where such protection is seen as especially important for public policy reasons. And by the way, that's true even in the early republic when, when defamation law was a lot broader than it is now. So just to give an example, as a general matter, you, you can't sue a witness for defaming you for something that the witness said in court, because the, the sense was that that would just chill witnesses too much. That if a witness says something that may be mistaken, not perjury, not even not a lie, but just an innocent error, then that person could be financially ruined because, because the person they said it about would, would sue them as a result, people wouldn't show up to court and wouldn't testify, or at least would try to kind of testify as little as possible. Another example is the fair report privilege, where when people are writing about what was said in court, could be a journalist, could be an ordinary citizen who says, you know, in this court opinion, the following things was said, or in this court filing, the following thing was said, that thing might be, might be mistaken, but, but the person who is reporting on it would have immunity not as a First Amendment matter, but just as a matter of these privileges that are internal to defamation law. So yeah, even in the absence of First Amendment, there would be protection for at least certain kinds of things. And of course the issue in defamation law is how do you deal with, with balancing two concerns, right? It's, if it were only about one thing, then it would be very easy have no defamation law. In fact, New York Times v Sullivan, three justices said, at least for public officials, when there are statements about them on matters of public concern, no defamation liability at all, even outright lies just too perilous to allow lawsuits over that. But even Justice Brennan writing for the majority wouldn't go that far. So on one hand, we don't wanna unduly chill people's coverage, either of public officials or public figures or ordinary citizens who somehow get involved in, in some debate, sometimes voluntarily, sometimes not. But on the other hand, we do want to chill people from saying false things that damage other, damage other people's reputations. So the question is always, how do you, how do you try to minimize the chill and free speech while at the same to providing some protection for reputation? How do you provide some protection for reputation while at the same time minimizing chill on free speech?

- Can I ask a question that relates to that? Oh, go ahead, Larissa.

- Right. I was just gonna say there's one other interest that comes into play too. So defamation is about protecting individual re reputation, which is linked in part to conceptions of individual dignity. But defamation law can play a role in keeping the public discourse honest. That defamation is supposed to partly serve to make sure public discourse has some necessary anchor in truth. And people would argue maybe it's lost sight of that, but that that falsehoods, that pollute public discourse and injure reputation at the same time don't, you know, deserve legal protection.

- So my ideal system to sort of get at Eugene's, the tension that Eugene raised would be to allow the cases, but only on nominal damages. This brings, this brings the, I know Larissa, you're, you're thinking about writing about damages right now. And so it's now that we've seen, we're in the era where these damage awards keep increasing, I wonder if that might be, you know, everyone's doing the kind of cal calculation of the estimated risk and as the losses, if we, as we all remember from the hand formula, as the losses get higher and higher than the expected value of the loss, you know, of, of what you might have to pay out gets higher too. And yeah. So what, what role does maybe capping damages play in all this?

- Yeah, so that's one of the things that we're looking at in the restatement and in the, in the, we don't, this is not part of the approved draft yet. This is literally in the earliest drafting stages, but we've talked about defamation has a weird kind of damages called presumed damages. Meaning damages that flow from the very nature of the statement itself. Like the jury can just say, if somebody says you're a child molester, that sounds like it's worth about $10 million, right? That it, that like damages are presumed to flow from that kind of statement. And unlike many torts where you have to prove damages with some degree of specificity in presumed damages, the jury just gets to kind of like, you know, vibes it. Well. So one of the things we have, we are looking at is aligning defamation law with other dignitary harm types of torts, like intentional infliction of emotional distress, and saying, presume that you can still have nominal damages, damages in name only to vindicate the interest that to kind of work as a statement that yes, you were defamed and it was false, what was said about you was false and it shouldn't have been said. But unless you can prove your damages with some kinds of evidence, then you shouldn't be able to get these windfall, you know, massive amounts of damages. And also it would eliminate things that you're seeing these days. I mean, people are suing and claiming billions of dollars in damages for defamation, which is kind of a,

- I mean, it, it's becoming the sort of thing where it would be cheaper to just kill the person than to actually lose a defamation alert to them.

- Yes, I'd rather go to jail than pay a billion dollars. So yeah, so, you know, criminal defamation has largely fallen by the wayside as, as deeming obsolete and outmoded, although it still exists in some places. But I mean, better that than just like a windfall billion dollars. One of the interesting things about AI is at least when AI preserves records, we may be able to tell how big the audience was, you know, for the defamatory statement. Yeah. Or how many people access it, how many people searched and saw the defamatory statement, which could give us a benchmark for deciding what the damages were. Hmm, hmm. It's an interesting idea, but yeah, I'm not a fan of presumed damages even when the presumption is rebuttal rebuttable.

- I wanna ask you one last question, and it also bears, although someone indirectly on this shift the internet from, from the traditional print model of defamation. So it used to be the defamation was a short, sharp shock. You wake up in the morning, you go out the front on, on your front porch, you see the local newspaper, it says something false about you, and that's, that's the damage to your reputation, and it's unlikely that they're gonna keep banging that drum. Whereas with the internet, it's the drip drip drip. You see there's some page which maybe have had relatively few readers so far, says something false about you. And, but then the problem is that it's, people are gonna keep coming back to it. And maybe so far there's been very little damage, but whenever somebody googles your name, they're gonna see that page. Maybe it's gonna end up feeding into AI algorithms that, that will then dutifully report what that page says. So this bears on a, a couple of questions. One is this old question of whether there should be injunctions in libel cases. It used to be that that was considered the quintessential unconstitutional prior restraint, but it was also pretty rare to even be able to get an injunction because how do you know that something is coming down the pike about you, especially in time for you to go to court. Whereas now, you know, you could say, look, something's been posted about me, I don't want money. I mean, I would like money, but I'm not gonna get any money from the defendant. I just wanna make it stop. So that's one area. And I think we have seen rightly or wrongly courts in a majority, I, by my count, over two thirds of the states actually issuing and often expressly allowing injunctions in libel cases, at least following a trial on the merits where the speech was found to be found to be false and defamatory. But another question that comes up is how we should measure things like mental state, like knowing falsehood and such for something that's, that's up and someone is keeping up. So let's say a reporter writes an article about me, and it's false, but it's an honest mistake on the part of the reporter. And let's assume that for this purpose, I'm say a limited purpose public figure. So I have to show knowing falsehood and, and I can't because, because the reporter credibly says, you know, I thought it was true at the time. And I say, okay, fine, so I can't sue you for having published this, but you still have this on your website, and now you do know that it's false and you are keeping it up in the face of this knowledge. So even if it wasn't a lie at the outset, it's essentially retroactively, in a sense, a lie. Now at least you are keeping it up is something you are doing with knowledge of the falsehood, or at least recklessness. Maybe, maybe you aren't sure that it's false, but I've given you enough evidence that you are aware that it's very likely false. If only I had, I had gotten the evidence, the reporter before the reporter published the article, then in that case I'd be able to sue. But I did, I I couldn't because I didn't know it was coming. So then, so then it's up. But I just want it taken down. My reading is that most courts say, Nope, we only look at the mental state as of the moment it was posted. And even if the reporter or site operator or somebody else, it doesn't matter if it's the media or not professional media or not, even if the, if now they're fully aware the statement is false, they have no obligation to take a duck. Not at all. There are some cases that do suggest some such obligation, but that's not the dominant view. So what do you think about these two questions that bear on this kind of new, new internet nature of defamation being an ongoing process rather than just an event that has happened? How you think, if at all, this should bear on the injunctions question and this question of whether there should be liability for keeping something up once you're aware that it's false?

- Okay, well I'm super interested to hear what Jane says on this too, but, so I think there's a problem anytime you create take down obligations before there's been an adjudication that, that it's defamatory and false.

- Oh yes, I agree.

- So, because it just, it is really a difficult thing to determine what's defamatory. And even the most obvious sounding defamatory statements can be turned into non defamatory opinions given context. So for the example, Dr. Smith is a murderer. Oh, why are you saying that? I'm saying that because I'm an abortion protester. I think abortion is murder and Dr. Smith performs abortions, right? That's clearly, I mean, assuming it's true, he performs abortions, then it becomes, this seemingly factual statement becomes opinion. And so it's very hard just anytime anybody sends you any kind of take down notice to say, oh, well I need to hop to, and so the law is, is understandably reluctant to impose obligations anytime anybody asks for take down. I do think we, we learn in our research for the restatement that the, the law like has developed some sensible responses to that. So for example, in the cases where a property owner has a bulletin board or a wall or, or something where somebody posted a defamatory statement, they're not responsible for that defamatory statement, even after they're notified of the defamatory statement, unless them leaving it up there suggests that they ratify or endorse the statement. And so, and that, that also, you know, we, we argued that that's the way, for example, libraries should operate because there's no case holding an American library responsible for refusing to take down a or to take out a specific book that they've been notified is defamatory. And you wouldn't want that, right? You wouldn't want the strategery that goes along with notice and consent based regimes with someone that doesn't have a real incentive in protecting expressive rights at the expense of them potentially being held liable. So I think it's, it's a hard issue because of those concerns, but I'm super interested to hear Jane's thoughts on it.

- Well, I don't know that I have any specific thoughts. I guess I kind of wanted to treat, I get the problem of the host who's trying to just stay clear and not be involved in the conversation. The bar owner who doesn't wanna paint over the bathroom door. Yeah, bathroom stall door. But, but as to the original content creator and the internet problem that Eugene laid out, I mean, would there be room to consider the failure to take down a post as if it were republication? I mean that's, I know it doesn't, that's not really what it is. But knowing that the internet is a, you know, if it's thought of as a con, you know, a continuous message as a service, then that might be one workaround that may make some sense

- If you have continuous restart of the statute of limitations, or how do you, how do you work that? Yeah.

- Well, but the statute of limitations I think is an important point, but it's a somewhat separate one. So statute of limitations, I think, which is in many states, just a year for, for defamation pretty short by the standards of other things. In I think most other states it's two years and no more. I think that makes a lot of sense in part because, you know, after a year or two, evidence may get lost, it may become harder to remember just what the conversations were harder to, to, to find the relevant records and such. But in a lot of these situations, the, the timeframes are really quite short. So just an example, yeah, let's say, let's say that I'm writing something about you and you call and leave a voicemail on my, on my phone, and you say, VALIC, look, the thing you're about to tell me to say about me, it's just false. It's case of mistaken identity. There's another ky this is what happened. And I get that message in time before I publish my article. If I ignore that, and let's assume the, the message lays out all the facts in a way that are, that I realize at the very least, it's very likely that I'm about to make a mistake. If I ignore that message at that point and just publish without that, without reference to that message, I am guilty of libel because I'm repeating a false factual assertion, not just a matter of opinion. And I know that it is very likely false, and I'm barreling on despite that. That's recklessness, that's so-called actual mail. But let's say I'm less assiduous about checking my voicemail. I don't check my voicemail until an hour after I post the article. At that point, I, I have the same mental state. I'm still aware an hour is the only difference in timing between the two of them. It's not like information would've gotten lost during that hour, but at least the way many courts say it, it sounds like I am now completely off the hook. I can keep my article on my site as the statement that I'm endorsing, even though I am now aware of, of this information. Because at the time I published it, I was unaware of the information. So I didn't, didn't have the, the requisite mental state wasn't reckless about the falsehood. Now I'm aware that it's very likely false, but it's already published. So, so my obligation is done. That just seems odd. I

- Mean, clearly it's odd, but I, I think it shows that they were making too strong of an analogy to the newspaper crawl an hour later with a newspaper that has gone to the presses or that has been distributed. It, the claw back is extremely expensive, whereas taking it off for the website is extremely cheap. Right. So I, I think this is a area where, where, you know, we, we should be careful with the analogies we make. We should design new rules where the, you know, where the, the thought behind the, the original rule doesn't make sense anymore. And you know,

- I think this ties into kind of like what you and I both talked about in our articles about AI and LLMs and, and defamation. If it's really easy to do targeted takedown of the specific ambulatory statement, why it, it's irresponsible not to. And should the liability rules follow the, you know, the bad, shouldn't they punish the bad behavior of the defendant? And in this case, in the case you described, there's an added benefit is as long as that's out there, it is polluting the public record, right? With a false hood, like a reputation saying falsehood. But it's also a false hood that that is bad for, you know, public discourse, public debate. And so aren't there added reasons, you know, within the realm of free expression to think that, that there's a benefit to, to put it, making an obligation of take down.

- Yeah, you've persuaded me.

- So

- I think

- It went the other way, but yeah,

- So Larissa, such a pleasure talking to you, partly because it's always a pleasure talking to you, but I think it's also such a pleasure talking to, to someone who totally knows what she's talking about and who's been living and breathing this. I mean, I think in your case, you've just been a, a media law scholar for so long, but you've been living and breathing it even before the reporter position. But now it's just so intense, both between the, between the, the reporter position and your work on the treatise that it's just very interesting for me and I hope was interesting for our listeners as well, Jane. Yeah, it's also great always talking to you. I'm sorry, sorry, I I focus so much on Larissa, but she's our guest, so Yes, exactly. So, so in any case, very much enjoyed this. And to, to our listeners, we will have another one of these episodes coming down the pike real soon. Now

- This podcast is a production of the Hoover Institution, where we generate and promote ideas advancing freedom. For more information about our work, to hear more of our podcasts or view our video content, please visit hoover.org.