Your Face, Your Voice, Your Self: Deepfakes, Biometrics, and the Law of Identity

Friday, October 16, 2026

Overview

Generative AI can now clone a person’s face or voice without their consent, while biometric systems turn those same traits into data used to identify and track people — raising the question of who controls identity once it becomes data. In this Berkeley Technology Law Journal symposium panel moderated by Sonia Katyal, right-of-publicity scholar Jennifer Rothman and online-abuse scholar Mary Anne Franks join SAG-AFTRA’s chief negotiator, Duncan Crabtree-Ireland, to debate whether existing law is up to the task, weighing the NO FAKES Act, the limits of consent, and Section 230 — and what each means for performers and ordinary people.

 

Instructor(s):
Duncan Crabtree-Ireland, National Executive Director and Chief Executive Officer, SAG-AFTRA
Mary Anne Franks, Eugene L. and Barbara A. Bernard Professor in Intellectual Property, Technology, and Civil Rights Law, GW Law
Jennifer Rothman, Nicholas F. Gallicchio Professor of Law, Penn Carey Law School
Sonia Katyal, Associate Dean, Faculty Development and Research, Roger J. Traynor Distinguished Professor of Law, Co-Director, BCLT, UC Berkeley Law

5 Key Questions Answered

1. Can existing law handle AI deepfakes, or are new laws needed?

Duncan Crabtree-Ireland argues that everyone who has tried to use the current patchwork of remedies has found it unworkable, that it gives platforms no real reason to remove abusive material, and that this is why he supports the NO FAKES Act. Jennifer Rothman frames it differently: every state already recognizes a right against unauthorized use of a person’s name, likeness, or voice, and those laws apply to digital replicas — though she later clarified that this framing does not mean new laws are unnecessary.

“I think the sort of patchwork quilt of various remedies… It’s a very thin patchwork quilt, and my evidence of that is the fact that anyone who has tried to use any existing legal remedy to address this kind of problem has found it to be entirely unacceptable. It’s… not a functional way to help provide any kind of limitation on people’s ability to abuse your image, likeness, and voice. It does not provide for any real motivation for platforms to cooperate in removing material, and that’s why we need things like the NO FAKES Act.”— Duncan Crabtree-Ireland

“In the aftermath of that [the rise of instantaneous photography], we did get change. We got legal change. We got adoption in every state of some version of either a common-law right against unauthorized uses of a person’s name, likeness, or voice, or a statutory version of that. Those have been around for a very long time, and they apply to digital replicas, right? Because you mentioned this. Now we can replicate people, and it is true that it’s easier and it works better. But as someone who also worked in the film industry, for decades we have been able to make really good versions of people, and now it’s just easier and more of a problem. But that doesn’t mean that suddenly this law against unauthorized uses doesn’t apply.”— Jennifer Rothman 

 2. What would the proposed NO FAKES Act do — and what would it leave out?

Rothman explains that the bill would create a federal digital replica right, but it is not a broad federal right of publicity. She warns that it largely ignores deception of the public: as long as the person depicted agreed, it allows — and may even encourage — replicas that mislead viewers.

“Drilling down into NO FAKES, the NO FAKES Act, for those who don’t know, is a proposed federal law that would create a digital replica right. It’s really important to just highlight at the outset what it does and doesn’t include. So, it is not a broad federal right of publicity.”

— Jennifer Rothman 

“Second is that it [the NO FAKES Act] is not concerned with some of the other things that we might be concerned about in the context of deepfakes and digital replicas, notably the deception of the public. So, it… not only allows but in some way incentivizes digital replicas and deepfakes that deceive the public as long as someone agreed to them, right?”

— Jennifer Rothman

 3. Why does SAG-AFTRA support the NO FAKES Act even though it is narrow?

Crabtree-Ireland argues that narrowness is the price of passage: broad AI legislation draws too much political opposition to survive, and perfection should not block real improvement. He stresses that the bill’s platform notice-and-takedown remedy is what ordinary people need, because they cannot spend years litigating against an unknown party — a lesson from his own experience being deepfaked.

“We also have to recognize NO FAKES is narrow. I agree with that assertion. The only way NO FAKES is going to get enacted is by it being narrow. I mean, look at what’s happened with every attempt to create omnibus AI-related legislation. You include too broad of a scope, and what happens is there are too many countervailing forces that then use their political power in our really messed-up political system to stop them from happening.”

— Duncan Crabtree-Ireland

“Regular people don’t want to spend their life pursuing litigation. I’m a lawyer, and I’m constantly encouraging people not to litigate things because I’ve seen how badly that can go for people who have to invest their energy and their money in pursuing litigation. So having a notice-and-takedown process, like a platform remedy, that is what most of our members and I think most people in general want. And I know, again, not to use my own example too much, but in my case, you know, my goal wasn’t to try and find a legal cause of action against this still unknown party and try to sue them for some money. My goal was get this [deepfake] taken down ASAP because I didn’t want more people continuing to think that I believe something I didn’t believe.”

— Duncan Crabtree-Ireland

 4. Is consent enough to make a digital replica acceptable?

Mary Anne Franks calls consent a floor, not a ceiling—uses without consent should be off the table, but consent alone is often not sufficient given exploitation and vulnerability. Crabtree-Ireland points to SAG-AFTRA’s contract standard — a reasonably specific description of intended use — which he says has pushed companies to come back for fresh consent rather than stretch an old one.

“Generally, what I’ve thought is consent is a floor, not a ceiling, so I certainly think that things done without consent should be off the table. Do I ever think that consent is necessarily sufficient? I think the answer in many cases is no, for reasons that you have pointed out. That exploitation, vulnerability, susceptibility to other interests, those things really do need to be taken into account.”

— Mary Anne Franks 

“We negotiated a reasonably specific description of intended use in 2023 as the standard for informed consent for digital replication under our contract with the studios and streamers, and so I mention that only to say we have now about three-plus years, almost three and a half years, of real-world experience using the standard of reasonably specific description of intended use. And like everything else, is it perfect? No, it is not perfect. But we have found it to be very functional in the sense that when somebody does enter into a license for a longer term, in our case, those are typically less than seven years because of the seven-year rule that’s applicable under California law. But whether it was five years, seven years, 10 years, a longer-term consent, what we have found already is that they are going back to secure additional consent later on in the future.”

— Duncan Crabtree-Ireland  

 5. Why does Section 230 keep coming up as a lever against deepfakes?

Rothman notes a circuit split over whether right-of-publicity claims fall within Section 230’s exception for intellectual property claims. She says a one-sentence fix placing them inside that exception would push platforms to take material down, as they already do for copyright. Franks names reforming Section 230 as her single legal fix, tying the industry’s reckless attitude to how the statute has been interpreted — while noting her view that AI outputs themselves are not a Section 230 issue.

“Section 230 is sort of an elephant in the room here. Instead of 49  to 98-page, very difficult-to-parse legislation that no one will understand, particularly ordinary people. One of the main reasons that companies don’t take things down is Section 230, and there is a circuit split on whether state publicity- and privacy-based appropriation claims fall within 230’s immunity for platforms or fall within an exception for IP-based claims. The Third Circuit has said that right-of-publicity claims fall in the IP exception… The Ninth Circuit has taken the opposite tack. One easy way that’s about a sentence, right? You could either strike Section 230, which isn’t likely to be very popular, but a one-sentence conclusion that right-of-publicity laws and appropriation claims fall within the intellectual property exception would provide an incentive for platforms to take things down. They take things down very expeditiously with copyright, which does have the DMCA takedown provision.”— Jennifer Rothman 

“The tiny thing that I would suggest, and I know it sounds outdated at this point, probably if it’s a legal fix, is actually to do something serious about Section 230. I personally believe this is a whole other conversation about how AI is not a Section 230 issue because Section 230 is supposed to be about user-generated content, not the actual outputs of these entities themselves. But I firmly believe that that attitude, that reckless attitude that we see from the industry, that they can do whatever they want and never have to clean up it later, comes from the way that they have interpreted Section 230 and the way that it’s been interpreted. And we need to do a very serious reset of that and say this is—you are going to be held accountable for your actions.”

— Mary Anne Franks

Summary generated from the recorded interview/transcript and reviewed by BCLT staff.

Keywords

NO FAKES Act • right of publicity • digital replicas • deepfakes • Section 230 • Illinois Biometric Information Privacy Act • Take It Down Act • notice and takedown • consent to digital replication • SAG-AFTRA AI protections in collective bargaining • does right of publicity law cover AI digital replicas • Section 230 intellectual property exception for right of publicity claims

B-CLE Recordings (CLE: Free) | Youtube Recordings (Non-CLE) | Agenda | Resources | Speaker Bio(s) & Contact Info | Summaries

Download the transcript here!

Interview/Transcript

This interview/transcript was based on a conversation from the 7th Annual BTLJ-BCLT Fall Race & Technology Symposium: Being Human in the Age of AI about Your Face, Your Voice, Your Self: Deepfakes, Biometrics, and the Law of Identity, on October 16, 2026, hosted by Berkeley Center for Law & Technology, UC Berkeley School of Law. The conversation was presented by Sonia Katyal, right-of-publicity scholar Jennifer Rothman and online-abuse scholar Mary Anne Franks, along with SAG-AFTRA’s chief negotiator, Duncan Crabtree-Ireland. 

Brandon Zhong  0:25  

Hello, I’m Brandon. I’m a member of the BTLJ [Berkeley Technology Law Journal] Symposium team. Thanks for watching Panel One. That was such a great discussion. Now I’m pleased to introduce to Panel Two: your face, your voice, yourself, deep fakes, biometrics, and the law of identity. Generative AI can now create convincing replicas of a person’s face or voice without that person’s consent or knowledge. Meanwhile, biometric systems can turn those same traits into data used to identify and track people. Existing laws protect different interests and do not always treat these harms in the same way. This panel therefore asks a central question: Who controls identity when identity becomes data? Our moderator today is Sonia Katyal, the Roger J. Traynor Distinguished Professor of Law and Co-Director of the Berkeley Center for Law and Technology [BCLT]. Our first panelist is Jennifer Rothman, the Nicholas F. Gallicchio Professor of Law at Penn Carey Law. Her scholarship focuses on the right of publicity and personality rights. Our second panelist is Duncan Crabtree-Ireland, the national executive director and chief negotiator of SAG-AFTRA [Screen Actors Guild–American Federation of Television and Radio Artists]. He leads the union’s technology and innovation work with a particular focus on AI and its impact on performers and the entertainment industry. Our final panelist is Mary Anne Franks, the Eugene L. and Barbara A. Bernard Professor in Intellectual Property, Technology, and Civil Rights Law at George Washington University Law School. Her work focuses on civil rights, free speech, and technology. Welcome.

 

Sonia Katyal  2:22  

Welcome everyone to the panel: Your face, your voice, yourself, deep fakes, biometrics, and the law of identity. I’m really, really excited to be moderating this panel. My name is Sonia Katyal. I’m a law professor here at Cal, and I work on areas relating to intellectual property [IP] and civil rights. And I’m just deeply thrilled to be here. So generative AI can now clone voices. It can fabricate likenesses. It can create convincing images of people doing things that they haven’t done, and at the same time, we also see biometric systems increasingly identify, surveil, and track people through their faces, their voices, and their bodies. And so, the point of this panel is actually to explore the sort of emerging and dynamic law of identity, including digital replica laws, the NO FAKES [Nurture Originals, Foster Art, and Keep Entertainment Safe] Act, biometric privacy statutes, facial recognition harms, and how it all ties into the right of publicity, and we’re really trying to focus on the question about who can control identity when identity has increasingly become more datafied in this current era.So, in this sort of process and in this moment that we’re living in, what we see is that the law has kind of historically treated identity as something stable, like one’s name, one’s image, one’s voice, one’s likeness, but generative AI, as we’re seeing, is really disrupting the stability with its ability to convince, to produce convincing versions of a person’s face or voice, often without their participation, their consent, even often without their knowledge, and so how do we struggle with these kinds of changes that technology has created, and particularly how do we struggle with all of these changes in a world where we’re seeing biometric systems turn these very same characteristics into data that can be used to identify and track people, so the law has given us tools to deal with this problem. We have the right of publicity, we have the right of privacy, we have copyright, we have First Amendment doctrines. But each of these protect different interests, and so a celebrity whose likeness is commercially exploited, a performer whose voice is cloned, an ordinary person whose face is scanned without their knowledge, we’re all experiencing different kinds of what we might see as sort of a violation of something more diffuse, something like perhaps the right to identity, and the law doesn’t necessarily treat all of these different interests in the same way.So the emerging fight over digital replicas has actually made this particularly visible. When we see the proposed NO FAKES Act, which would create a federal intellectual property right in one’s voice and likeness and establish liability for unauthorized AI-generated replicas. So, in this process, in trying to empower individuals to take action against bad actors who might knowingly create, distribute, or profit from these unauthorized digital replicas, this new legislation actually raises a really fundamental question about how these new areas of law can interact or should interact with older areas of law, like speech, parody, news reporting, and other kinds of expressive uses. And so, we have all of this sort of circulating at the congressional and also at the cultural level, and at the same time, we also see the emergence of other kinds of statutes, like the Illinois Biometric Information Privacy Act [BIPA], which has started to treat biometric information as deserving of kind of a special form of legal protection.So the question that I really want all of us to focus on today is who controls identity when identity has become more like data than anything else? When our faces, voices, our likenesses can be collected, reproduced, and used without our involvement, how do we decide who has the power to govern that identity? Is it the individual? Is it the companies using the technology or collecting the data? Is it the law, or is it the legislature? And so, in order to kind of unpack some of these questions, we have an incredible panel assigned for you today. We have Professor Jennifer Rothman, Duncan Crabtree-Ireland, and Mary Anne Franks, who are all going to speak on these topics. Let me just introduce each of them very briefly, even though their CVs are incredibly long and distinguished. With us, we have Jennifer Rothman, who is the Nicholas F. Gallicchio Professor of Law at Penn Carey Law School, globally recognized for her scholarship in the field of intellectual property, privacy, entertainment, and media, and is actually the leading expert on the right of publicity and personality rights. Wrote a fantastic book that I all recommend all of you to read, called The Right of Publicity: Privacy Reimagined for a Public World. We with us we also have Duncan Crabtree-Ireland, who is the national executive director and chief negotiator of SAG-AFTRA, led the union’s technology and innovation team with a particular focus on current issues like artificial intelligence and its impact on SAG-AFTRA members and the media and entertainment industry generally. Super honored to have you with us. And last but definitely not least, we have Mary Anne Franks, who is the Eugene L. and Barbara Bernard Professor in Intellectual Property, Technology, and civil rights law at George Washington University Law School. She is an internationally recognized expert on the intersection of civil rights, free speech, and technology, and also serves as president and legislative and tech policy director of the Cyber Civil Rights Initiative [CCRI], which is a nonprofit organization that is dedicated to combating online abuse and discrimination.So I want to thank all of our incredible panelists for being here. Each of you are such an inspiration in your own particular field. It’s okay. What I’d like to do is actually start off with Jennifer Rothman, Professor Rothman, your work on the right of publicity has examined the legal interests that people have in their names, their likenesses, and other aspects of their identities. So, if the right of publicity protects individuals from having their identities used without permission, can I ask you to help us frame how we should think about that right when AI technology makes it possible to create entirely new uses of someone’s identity? So, what exactly should the law be protecting, and how do we think about these tensions? Right between the economic value of identity versus the individual’s control over identity, or what other kinds of interest might play into this equation? Could you say a little bit about that?

 

Jennifer Rothman  10:13  

Sure. Thank you so much, Sonia, for moderating and for the lovely introductions. And I so appreciate our time together, and have tremendous respect and appreciation for all the work that the other panelists have done in this area. I and I wish we were all were in a collective online space. I wish we were in a physical space together. And I know this will be delayed in its broadcast, but I hope people will still feel they’re part of the conversation.

I think I want to take a step back from sort of where that question started us to just remind people that whenever there’s some sort of technological change or leap, people wonder if the law is up to the task, and sometimes we rush in and we make changes to the law about technology that then becomes obsolete before the laws even go into effect. Sometimes we forget what laws are already on the books. Sometimes the changes reveal whether they’re technology based or market-based or societal, that there are gaps in the law that we need to do something about. So I want to sort of set the stage to think about that and to use that as a metric, which is I like to think, and this is maybe like in the med field more than the law field, right? But the sort of do no harm. So when we think about legislating new, you know, new laws in this space, are we making things better? Are we making them worse? Are we making them more confusing or complicated? And so, I want to start with that as we sort of drill a little bit more into this topic. And I don’t necessarily think that this technology alters how we’ve been thinking about personality rights in this country. The right of publicity is a dominant one. It’s not the only one. I think it’s very important to also highlight that the right of privacy and the right of publicity are essentially one and the same. Those terms have been used interchangeably for over 100 years, and came out of another technological moment, which was the rise of instantaneous photography. And people wondered how can we stop? How can we stop people from taking our picture when we don’t want our picture taken and we’re just walking on the street? How can we stop people from publishing images of us in newspapers or talking about us in newspapers during the era of what some call yellow journalism? And also, how can we stop uses of our names and likenesses without our permission in advertising and on products, so we see a lot of the same conversations we’re having now, but happening in the mid to late 1800s, and so I think that’s just an important thing to remember. And in the aftermath of that, we did get change. We got legal change. We got adoption in every state of some version of either a common law right against unauthorized uses of a person’s name, likeness, or voice, or a statutory version of that, right? And those have been around for a very long time, and they apply to digital replicas, right? Because you mentioned this. Now we can replicate people, and it is true that it’s easier and it works better. But as someone who also worked in the film industry, I’m like for decades we have been able to make really good versions of people, and now it’s just easier and more of a problem. But that doesn’t mean that suddenly this law against unauthorized uses doesn’t apply. So I think that’s important to remember. There are a host of other state laws, and we can talk about those separately that apply. Some federal laws, in terms of particularly for people who are well known, like trademark and unfair competition law, both the state and federal level. In terms of a lot of Mary Anne’s work, there’s at the state level a lot of laws now about intimate images. There’s also defamation laws, false light laws, emotional distress claims. There’s all sorts of law that already exists in the space.

 

Jennifer Rothman  14:32  

And so, what I think that this moment of technological change has done is to shine a light on some of these issues when most people weren’t necessarily paying attention to them, and I, for more than a decade, or maybe approaching two decades at this point, have been very concerned about people losing control over their identities and the ability to reanimate people, to recreate using prior record. People’s voices and images, and so to me, this is a great opportunity to reflect on what’s working and what’s not. And I’m sure we’ll have more opportunity to talk about NO FAKES, which is not law. It is a draft bill that’s been introduced a few times and has gone through a few changes, most notably in June of this year, an entirely sort of new version of it was introduced. But I feel like I’ve already occupied a lot of time, so I’m happy to talk more about that. But I just want to sort of start with that sort of taking a step back view.

 

Sonia Katyal  15:36  

Thank you so much. This super helpful. I’d like to actually turn to Duncan. So, Duncan, you led SAG-AFTRA’s technology and innovation work. You had a you have a really interesting particular focus on AI and its impact on the media and entertainment industry. I would love for you to just reflect on what Jennifer has kind of set out for us to think about, but also in terms of your own work, what your work might have revealed about how existing legal frameworks have fallen short, or if they’ve fallen short when it comes to protecting people’s identities.

 

Duncan Crabtree-Ireland  16:19  

Yeah, thanks so much, and like Professor Rothman, like Jennifer, I’m so thrilled to be on this panel and a huge admirer of my co-panelists and of you as well, Sonia. So thank you so much. And let me just say that I think this is a really critical topic. And Jennifer and I have had this conversation a couple of times, and I don’t think we quite agree on all of the sort of implications, but I think we do start from a common foundation, and that is that you know this technology. This is not completely new technology, right? There have been capabilities in the past before the advent of generative AI tools that are available to the general public, and it has been used in, frankly, isolated cases in the further past, and as it works its way towards the future more and more frequently. For our members, this became a very clear existential threat, and I think sooner than it did for most people, because for most people, you don’t necessarily make your living from the use of your image likeness and voice, and from controlling the use of the your image likeness and voice, and many of our members do that. So I think it focused our attention on it earlier than most people. But frankly, I think everyone should be paying attention to this issue now. And I’ll just use myself as an example. And I think one or several of you may have heard me use this example before, but I myself was deepfaked, and this was in six months ago. This was in 2023, right after we had our big strike. You might remember there was a strike with the Writers Guild and SAG-AFTRA. We basically shut down the entertainment industry for about five months total between the two of us. And as we were putting out a vote to all of our members on the deal that we reached with the industry, which included a lot of rules about digital replication and other AI-related pieces. Someone decided to make a deepfake video of me using just publicly available material they scraped off of Instagram or the internet, and that video wasn’t just for fun. That video was taking me, my face, and my voice, and using it to try and convince people to vote against the agreement that I myself had spent a year engineering and negotiating and fighting for, and so I say this both from a perspective as a professional and a lawyer working in this field, but also as someone who has felt what it feels like to have someone take your face and your voice and use it to actually disseminate ideas that are anathema to you. I, you know, that is there is a feeling that engenders in someone who experiences that that I wouldn’t wish on anyone else because it is a it is probably second to the kind of abuse that many of our members suffer from deep fake pornography and sexually explicit material. Say probably second to that, it is feels very violative to have people think that you believe something that’s the opposite of what you believe because someone’s abused that technology. Where I think I differ from Jennifer is I don’t think it’s just a difference of degree, and the reason I don’t think it’s just a difference of degree is because the capability of this technology, at least as far as having any kind of widespread availability of it, is entirely different. That’s why now we have school children who are being abused by deepfake technology. We have teachers. We have things like the Joe Biden robo call that is impacting our political system. The impact of all of this is more than just it’s a little worse than it used to be. The other thing that I think you know I take no issue, of course, with Jennifer’s analysis of all the different legal remedies that are available. That’s absolutely true. But I think just as the sort of I’ve never been a fan of the patchwork quilt theory of moral rights in intellectual property law in the United States.

 

Duncan Crabtree-Ireland  20:08  

I think it’s a fig leaf attempting to cover up our inability to provide the kinds of moral rights for intellectual property creation that we see in other parts of the world, I think the sort of patchwork quilt of various remedies-it’s-it’s a very thin patchwork quilt, and my evidence of that is the fact that anyone who has tried to use any existing legal remedy to address this kind of problem has found it to be entirely unacceptable. It’s not a functional way to help provide any kind of limitation on people’s ability to abuse your image, likeness, and voice. It does not provide for any real motivation for platforms to cooperate in removing material, and that’s why we need things like the NO FAKES Act.

I think there’s economic elements to it, and I’m sure we can talk about those as well. But if we step back for a second from the economics and focus primarily on the question of consent and the ability to control the use of your face, your voice, and your image, then I think we are really falling short without additional tools. And I myself know because when we tried to get that video of me taken down from Instagram-it took forever to get it taken down. It only got taken down because Instagram voluntarily ultimately chose to do so. Tens of thousands of people saw it before it was taken down. We still don’t know what impact it had on the vote that was going on amongst our membership at the time, and I just think that’s just one tiny example. I mean, I could reel off a whole laundry list of them. The last thing I just want to say is, you know, I recognize that you know there is always a tension between intellectual property rights, freedom of expression, and I think we have to find the right balance between those things. And the example that was used of you know photography in public is a good one because there are so many people who incorrectly think that they have the right to just stop people from using a photograph of them that they took in a public place, right, for non-commercial purposes, and generally that is not true. But I do really want to draw a distinction between the scenario where someone takes a photograph of something that is actually happening, that you are actually participating in, that is happening in a public place, and somebody using AI technology to generate photorealistic images or outputs of you doing things, saying things, participating things, being in places that you were not at, did not say, and did not do, and then having you now become the person who has to defend that didn’t happen. That isn’t real, etc. I think it fundamentally impacts our ability to have a public discourse for people to control what people are hearing from them. And I feel that it constrains my freedom of speech to have other people be who perhaps might even have a bigger platform than me to put words in my mouth and disseminate them. I mean, I think this is a new kind of challenge that we really have to face. And while NO FAKES or other tools may not be absolutely perfect, I think we really, really need them. And not just for my members and for high-profile people or mid-profile people. All of us need them so that we can live in a society where we know what’s true and what’s fake.

 

Jennifer Rothman  23:19  

I would just since I was my views were assessed in a certain way. Without belaboring it, would just like to clarify that I think what I was doing was framing, and I think this is a problem of great magnitude, and that doesn’t mean we shouldn’t have new laws, or that the current laws are not enough. Right? It doesn’t answer either of those questions, and so I think there are some areas with which we disagree, which I hope we’ll get to talk about more and the sort of the nitty gritty of NO FAKES. I think there’s certainly a way in which federal laws in this place could help. Uniformity could help, although I note that SAG recently opposed efforts to have a uniform law in this space, and which maybe we can talk about more. But I think it’s very important to think about the large constellation of people involved, right? Not just performers, not just the wealthy, but everyone. And so, when we set up regimes, and maybe Mary Anne will talk about this, we want to make sure that the famous and the ordinary, those with power and those without, are able to control their identities and not just control them today, but control them in the future. And that we also don’t forget that the public is also a party here, not only from the perspective of free speech, which is very important, but also from the perspective of potentially being deceived by what I’ve called elsewhere deceptively authorized deep fakes. So, I hope we’ll have a chance to talk about that.

 

Sonia Katyal  24:54  

Great, thank you so much, Jennifer. I would love to hear from Mary Anne. Because your work, I think, has been really trailblazing in examining exactly how technology can facilitate abuse and discrimination. And so, I would love to hear your thoughts about kind of what changes when the unauthorized use of someone’s identity isn’t so much about exploitation, but about some of these other issues that both Jennifer and Duncan have talked about-sort of, you know, representations that people might disagree with, or humiliation, intimidation, misrepresentation. Do you think that the harms that we’ve all just spoken about are adequately addressed by existing identity and publicity laws. And where do you feel like we could go in the future?

 

Mary Anne Franks  25:51  

Yeah, thank you, Sonia. And I feel like it was so helpful to hear both your original framing of this conversation, and then to hear Jennifer and Duncan give their perspectives. Both I’m familiar with all three of you and your work to some extent, and so it’s so interesting to hear the insights that I think you know. As I’m listening, I’m thinking I not only agree with that, I also think that all of these pieces, even though there probably are some tensions and some areas of disagreement, that I think there’s so much commonality in the impulse at all, I think you know each of us has to say this is a new problem and an old problem at the same time. And I guess what I would add to that is to say, and the other thing we should learn from our history is it is because we have not adequately recognized those things that we keep finding ourselves in these really dystopian moments where technology has raced ahead. We are completely caught flat-footed. The law is racing to catch up. And even if the law gets there to some extent, at some point the damage is largely done. So even if we’re talking about successful civil litigation or even in some cases criminal penalties, the harm is usually not something that can be addressed, at least in any kind of truly meaningful way by those things, because these are unanswerable harms in some fashion that really can’t be compensated after the fact, and that I think that’s speaking to part of what you were asking. There are certain types of commercial, more quantifiable harms that we can think of that I think right of publicity does get to, copyright can get to the whole concept of intellectual property as opposed to let’s say a dignity framework or an autonomy framework. Not to say that IP is in contrast or thinks that those things aren’t important, but it’s the property framework is a very unusual shouldn’t say unusual. It’s it is a choice, right, to think of things as property that is both useful and limiting, I think.

So I guess I’m echoing to some extent the insights from before that it’s not just that the technologies have existed for some time. I would go back even farther to say we’re really talking at least in part about human impulses to treat other people in certain ways. That is, we don’t if we had a society where people actually thought that they needed to respect other people’s identities, that you didn’t want to take them out of context, that you didn’t want to capture their image or their speech at a time when they did not feel comfortable with you doing so. If we really felt that at a minimum we would not manipulate someone’s face or their voice to say something or do something that they had not done, then it wouldn’t matter to some extent. If we had all the tools to do it, we just wouldn’t do it. We it would be like you know many of us are fully aware, of course, that we have knives and guns all around us. We don’t pick them up and use them because most of us, most of the time, would not want to do that to another person. So we have to ask this question of why do people want to do this? We can call that a lot of different things. We can call it the instrumentalization of other people. Be a little catchier these days and say it’s partly this what is sometimes frivolously referred to I think as main character energy. But the idea that you are you, the person holding the technology, is the person who matters, and everyone around you is a prop or a form of some kind of raw resource for you to use in producing yourself and your story, and all these tools that help us. Those people that I think maybe throughout every period in society may have always been those kinds of, for lack of a better word, sort of puppeteers that have wanted to treat other people like this, and all the there may have always been a part of the society that would never want to do that to someone, and then a kind of squishy middle.

 

Mary Anne Franks  29:26  

Which, if you make it easy for them and you validate them for doing it, and there’s no downside to doing, might join in and become the kind of people who want to contribute to that kind of instrumentalization. And I think that’s what we’re seeing with all of these new tools. Duncan said it. Jennifer has also said it that tools make things easier. So it’s one. It’s not that it’s brand new, but if there’s hardly any effort that I have to make to exploit somebody, will I consider now doing it? Especially if it gets me points and some kind of social capital or actual capital. So I’m thinking about this from the perspective of what I’ve. I feel a little bit like. I forget. Like I don’t remember how many years it’s been to talk about this, but I think it’s going on more than a decade. But my very first law review article was called Unwilling Avatars, and it wasn’t about deepfakes because those didn’t exist, and it wasn’t really about AI. It was about women in particular and the way that they were being represented online, whether through message boards or through actual images, and at the time, really crude photoshops. You know, you know, manipulated images of, in particular, women who were successful bloggers or people who got into elite law schools. And so, some of you may remember this. I know for the younger generations won’t remember these phases where every time there was a new virtual environment or a new tool, it was used immediately to sexually objectify and to demean women and girls. And you can look all the way back to the 1993 Rape in Cyberspace article by Julian Dibbell, where he talks about this completely text-based game where one avatar in the text of the game rapes another avatar, and the description of the victim here saying that’s it felt like an actual assault because that person took over the character of someone else and made them do things that they didn’t want to do, and now you hear that in the virtual reality space, the virtual gaming space where assaults in virtual reality, according to the people who’ve experienced them, feel as real as a actual physical assault, and you have neurological studies that are saying what you experience in virtual reality is not like watching something; it’s like a memory of doing something. So you have all these iterations of these technologies, but they’re coming from somewhere. They don’t come out of nowhere. They come out of that impulse that may not have been solely about trying to exploit or to ventriloquize, but certainly were either unconcerned with the dangers of that, or maybe thought that’s a benefit, right? So I guess the bigger point that I would make here is that we those are the reasons where we keep finding ourselves in these positions is because we don’t listen to the people who are saying this impulse, right? Whatever tool is being used, but the impulse behind it is something we actually is the law’s job to try to prohibit because this is only going to get worse. So I think about it from yes, not just the commercial aspect because as much as I do agree with Duncan that there are very particular harms that are going to affect people whose business it is, whose livelihoods have to do with their voice and their image, and the public scale of their manipulation is probably going to be so much wider than your average person. There’s also sometimes a kind of corresponding ability to actually prevail upon your rights. That if somebody’s going to actually get their rights vindicated, it’s going to be maybe a Scarlett Johansson as opposed to someone, a school teacher, right? And to the extent that it’s not just about the scale of who sees the manipulated image or the impact it has in that sense, but it depends on who it is in your life that matters, right?

 

Mary Anne Franks  33:00  

If someone takes even a let’s not even say a photo manipulated image, but what we deal with a lot at the Cyber Civil Rights Initiative, which is an actual unmanipulated but private intimate image, and shows it to your boss or shows it to your child, or that is that impact of that audience, even if it’s one person, can be devastating. So a very long sort of circuitous route to your question, which is, do we have the right tools right now? Do we have? Is the law cut out to respond to any of this? I’d say absolutely not. And it’s not just because we’re not catching up to technology, but it’s because just to put maybe part maybe too fine a point on it, but for these purposes, that if we think about women and girls as what I’ve said, the canaries in the coal mine for technological exploitation. We keep not listening to their experiences because we think we, the society that’s not paying attention, thinks well, this doesn’t become real until it affects someone who matters, right? And the fact that it doesn’t ever seem to occur to people that women and girls do matter on their own, but also whatever gets tested right on women and girls will happen to everybody else, right? So when many of us were trying to raise the alarm about deepfake porn in 2017, we’re dismissed because it wasn’t having the political impact that people thought it might. They didn’t really see a lot of political deepfakes, so there were all these op eds written about how oh it’s an overblown concern because nothing real is happening, just 1000s, possibly millions of women and girls being exploited with having their images being used in this way. So that I think that we really do have to get at that question of what the law is supposed to do about human behavior, right? How it is that we can actually deter really destructive human behavior, and that includes using these tools, but also creating these tools to begin with, and really start getting creative-if that’s the right word for it-to think about how you go after that impulse and that ability to do these things. I’ve said before in a couple of different contexts that we should be looking more at things like the counterfeiting context, not just because counterfeiting and the kind of digital manipulation are similar in some ways because they degrade our currency, like our you know sort of social currency, et cetera. But also because you know federal law on counterfeiting says you can’t have the tools to counterfeit currency. Not just that you can’t do the actual counterfeiting, and that does raise freedom of expression concerns. What if you really want to counterfeit $1 bill for your art? I mean, there are real trade-offs here, but we don’t assume right that in the really important context of counterfeit currency that that means that we should let people counterfeit money and let’s just hope that they do it responsibly. So, when are we going to take this as seriously as we take property, as we take publicity, and that kind of larger sense of powerful people or people with meaningful harms that we recognize.

 

Sonia Katyal  35:43  

Yeah, I mean, I often think back to your incredibly prescient work on pornography in 2017, and how I think you really did sort of highlight for us the implications of what was to come and what we’re now facing today. So thank you for always having your eye on sort of the next generation of issues that we need to be thinking about. I want to actually turn to this deeper question that each of you have actually raised, which is it sort of I want to ask this question in kind of a general sense, but also a specific sense about the role of law in this process. So how do each of you view sort of the purpose of law in this process? You know, can we use the law to address some of the structural harms that Mary Anne has talked about, or you know, to put a more specific point on it. I would love to hear each of your thoughts about the NO FAKES Act, which, as I understand it, creates a federal right to control unauthorized digital replicas. Partly, I love to hear how, from you, Jennifer, in particular, like how you see it meshing with the sort of right of publicity law, but also for each of you to kind of get into the questions of scope in terms of how you view the extent to which this will protect everyone versus the sort of high-profile celebrities versus the kind of everyday individuals, all of whom it seems are implicated by these kinds of issues. So maybe we’ll start with Jennifer, and then Duncan. I’d love to hear your thoughts about the act, and then we can hear from Mary Anne. Does that sound okay?

 

Jennifer Rothman  37:37  

Sure. I’d love to those are really two separate questions. So I’m going to start with the law question, and I think there are lots of ways the law could be helpful. It needs to set a foundation of what’s acceptable, what’s not, provide a basis for civil and, in some instances, criminal liability for bad behavior. I think Duncan pointed out it’s important that platforms have an incentive to take infringing material down with appropriate speech protections in place. So, with that said, though the law has limits, so there are already laws in place that suggest that some of these things can be taken down. And Duncan’s absolutely right; the platforms have not done a good job of taking down material, even when they’re on notice. But even if we had the very best laws in place, the technology presents some challenges in terms of sort of a whack-a-mole problem. And Mary Anne knows this very well from her work, right? That you can get something taken down from one place, and then it could appear somewhere else. So I think we need partners with the technology companies in terms of design, in terms of ways of getting picking up material, of restricting material, but with you know protections also for free speech, things that are inspired by other style use of style, all of that.

Drilling down into NO FAKES, the NO FAKES Act, for those who don’t know, is a proposed federal law that would create a digital replica right. It’s really important to just highlight at the outset what it does and doesn’t include. So it is not a broad federal right of publicity. So there are lots of uses of right real images, explicit images, collage, all sorts of technology that’s new, not newly computer generated that would fall completely outside of the purview of This law, it also is people need to know that it’s on the backdrop of underlying state laws that problematically leave open the door for other people to own another human being’s name, likeness. Voice, and so that because it is not a problem that’s really fixed in NO FAKES or addressed or comprehensive, is a sort of dangerous foundation on which I think to legislate. Which isn’t to say that we shouldn’t, but we should be doing so with great awareness of that danger. And the NO FAKES Act has a preemption provision of state law, but it is somewhat difficult to parse and might take up all of our time together. The entire draft bill, I think, currently is over 49 pages, and that’s not including actually that it’s 98 pages because the first 49 are like struck out and then replaced with an amendment from June, so a few sort of things to keep in mind. One, NO FAKES was originally put forth by the recording industry, and so even though it applies to everyone, I think its name we were suggested by the journal to be careful about using acronyms. So we’re saying NO FAKES, but what it stands for is nurture originals, foster art, and keep entertainment safe. Right. So it has a very entertainment industry based focus, and that’s how it was originally drafted, even though it applies to digital replicas of everyone, and so we have to think about. And this is sort of like how the sausage is made, right on Capitol Hill. We start with one thing, people draft it, things get added, but it would apply to everyone. So in spite of that, there’s some specific provisions for those who are recording artists. There’s some carve-outs in the sound in the music industry where the record label would have more control over its recording artist than maybe others would have. SAG-AFTRA has gotten a carve-out for collective bargaining agreements, which suggests that some of the actual protections in NO FAKES wouldn’t apply if collective bargaining applies in that instance, which could be better or could be worse than what’s in the statute right now. It’s maybe it’s better, but one day maybe it won’t be. We don’t know. And then some of the then the sort of the second thing is right. So it’s focused largely on entertainment. Second is that it is not concerned with some of the other things that we might be concerned about in the context of deep fakes and digital replicas, notably the deception of the public.

 

Jennifer Rothman  42:36  

So it not only allows but in some way incentivizes digital replicas and deep fakes that deceive the public as long as someone agreed to them, right? So and authorized them. And the third thing I want to highlight is about this issue of control. So I mentioned that we have this underlying law that may allow others to actually own and control another person’s name, voice, and likeness. And I was very heartened to see finally in the June version that some language that allowed authorized representatives to make these agreements was I don’t want to say struck because that oversells it because it still allows people to have licenses to living people’s names likenesses and voices but suggested that any license for a digital replica would need to be signed by the underlying individual. Although the language is a little fuzzy there, but that was a major shift just this year. I think it makes it less chilling potential legislation that it was in May of this year, but still would largely allow an authorized representative to negotiate the terms, and the terms of the licenses are 10 years for a digital replica. And the bill itself says that anything can be done with the digital replica that was created during the 10 years thereafter. So it’s not really clear what the guardrails are on how that can be reused. 10 years is a really long time. There’s a suggestion that that would need to be a reasonably specific description of what the digital replica is during the 10 years. But as you can imagine, it’s very difficult to have a reasonably specific description of what a digital replica would do over a 10-year period, and so it’s likely to be a lot of wiggle room there, so that someone has signed on to something, they may not know the boundaries of how it’s going to be used, and will have signed out over all of their rights to sort of object to that. And fitting in with my second point about deceiving the public, like the public may not even know these are inauthentic performances. The person themselves will not have been able to have approved them, and so I think this raises a host of problems to the extent that we’re concerned about things like dignity, autonomy, common understanding of reality. That I would like to see. This legislation, if it proceeds to take a little more seriously, and that’s especially true because some of the one-off election deception laws and intimate image laws are being struck down as unconstitutional, understandably because they’re underinclusive. So, although this NO FAKES bill started as being about the entertainment industry, it applies to everyone, and it is an opportunity to protect everyone from having their digital replicas used without permission, and to get platforms to take some things down. So I think there’s some really good things in the draft, and also some very worrisome things in the draft, and also a lot of things that are very unclear in the draft.

 

Sonia Katyal  45:42  

Wonderful. Thank you so much for that incredibly insightful summary of the legislation. I’d love to hear Duncan’s reactions, and maybe just speaking from the position that you’re in, Duncan, of representing your clients and the concerns that are raised by the actors that you represent, what are your thoughts?

 

Duncan Crabtree-Ireland  46:04  

Sure, and let me just start out by saying, you know, I know people when they hear of SAG-AFTRA or think of Screen Actors Guild, a lot of times they think of our celebrity and high profile members. But I want to just remind everybody, we have 160,000 members. A very small percentage of them are high profile celebrities. A large percentage of them, are just regular working people who are trying to make a living, and we also represent broadcast journalists, recording artists, puppeteers. Mary Anne, I thought that was ironic when you were mentioning puppeteers. We represent the human puppeteers of puppets, among many, many others. And I say that to just point out that our interest in this is not driven solely by the interests of high-profile individuals, and I think the points have been made about some of them have the kind of resources to pursue these things where it bothers them in ways that any normal person really doesn’t. And I just want to make clear that for me, and I think for us, the way we look at this is how can a working actor, how can a news reporter we represent on a local TV station, you know, somewhere in this country, how can you know a recording artist who’s trying to get their first album out there, how can we help protect them? And I think that protection really aligns with what regular people, working people around the country need as well, and so I just wanted to say that I think for me, in with both of your questions, there is a through point, which is we cannot let perfection prevent us from achieving some kind of improvement here, and I think this is something that labor unions we are very accustomed to this because it is almost never do you go into a collective bargaining negotiation and walk out with 100% of everything that you want. Typically, you go in and you fight hard and you achieve the best possible result, and then what do you have to do? You have to come back again in X number of years, normally three in our case, but sometimes other links, and then you have to improve upon things and fight for things. If you insist upon everything being perfect, or you just don’t do it, then what you end up with is nothing. You end up with no improvement. You end up with a status quo that highly benefits technology companies, that highly benefits those who platforms and those who want to abuse them. So I, you know, I think the points that Jennifer has made, a number of them, I would agree with, and I think we have seen improvements in the draft of NO FAKES over time. But we also have to recognize NO FAKES is narrow. I agree with that assertion. The only way NO FAKES is going to get enacted is by it being narrow. I mean, look at what’s happened with every attempt to create omnibus AI-related legislation. You include too broad of a scope, and what happens is there are too many countervailing forces that then use their political power in our really messed up political system to stop them from happening. And so, from my perspective, NO FAKES is targeting a specific problem, and that specific problem is the one that we’ve talked about thus far in this panel. And I agree, deception of the public, other. I mean, those are major problems that we are also concerned about, and this may help a little bit in that direction because it gives the people whose image, likeness, and voice is being used to deceive the public some additional remedies, but it’s not meant to be a perfect solution to that, and it’s not even meant to be a perfect solution to this. It’s meant to be a huge improvement, and I think that it will deliver on that. I also really strongly believe that platform remedies are so important. They’re important to our members, and they’re important to regular people because the reality is regular people aren’t going to go pursue multi-year litigation against some unknown party in hopes that they’re going to achieve some kind of economic victory at the end. That is a theoretical possibility that has no functional value in the real world because regular people don’t have access to those kind of resources.

 

Duncan Crabtree-Ireland  50:05  

Regular people don’t want to spend their life pursuing litigation. I’m a lawyer, and I’m constantly encouraging people not to litigate things because I’ve seen how badly that can go for people who have to invest their energy and their money in pursuing litigation. So having a notice and takedown process, like a platform remedy, that is what most of our members and I think most people in general want. And I know again not to use my own example too much, but in my case, you know, my goal wasn’t to try and find a legal cause of action against this still unknown party and try to sue them for some money. My goal was get this taken down ASAP because I didn’t want more people continuing to think that I believe something I didn’t believe. I didn’t want my image used that way, right? And so I do think that when we focus on those key priorities, that this would be meaningful change that would give people remedies they don’t have now that they can actually use in the real world, and I think that is a critical component. And you know I commend Jennifer and probably Mary Anne. I’m less familiar with your work in the specifics of the NO FAKES area, but Jennifer and I have talked. We’ve been on panels together before, and we’ve exchanged our views directly. I you know I appreciate people pushing back on this and helping us make improvements, but what I also want to do is make sure that it actually does get enacted because if it doesn’t, then it’s really unhelpful. And the last point I just want to say is, as far as uniformity goes, I think there is value in uniformity if it’s good uniformity, you know. And you know, SAG-AFTRA’s point of view on this has been: we’re willing to support limited preemption. We’re willing to support the work that Jennifer and others have been doing at the you know Uniform State Law Commission on right of publicity, but we don’t want to see stronger protections weakened solely in the name of uniformity. So there’s a again a balance that has to be struck there as well, but you know I really hope and believe that we will see either the current or some future version of the NO FAKES Act enacted, and we really really need it. We needed it a year. We needed it three years ago. Sorry, I know I said this last thing, but the last thing I really want to say is because Jennifer mentioned reasonably specific description of intended use, and I realize I’m zooming in on one particular thing here. But the just for those who might not know where that comes from is that comes from our collective bargaining agreement. We negotiated reasonably specific description of intended use in 2023 as the standard for informed consent for digital replication under our contract with the studios and streamers, and so I mention that only to say we have now about three plus years, almost three and a half years of real world experience using the standard of reasonably specific description of intended use. And like everything else, is it perfect? No, it is not perfect. But we have found it to be very functional in the sense that when somebody does enter into a license for a longer term, in our case those are typically less than seven years because of the seven-year rule that’s applicable under California law. But whether it was five years, seven years, 10 years, a longer-term consent, what we have found already is that they are going back to secure additional consent later on in the future.

 

Duncan Crabtree-Ireland  53:25  

Sometimes paying additional money to get that consent because they don’t wish to take the risk that the consent that they obtained in the first place covers what it is that they’re going to do now because it wasn’t reasonably specific to that use, and so one way of looking at it as wiggle room, the way I look at it, and the way at least in my experience it’s played out in the real world, is it actually is sort of the opposite of wiggle room because those who are complying with the law in the first place are saying to themselves, or in this case, the contract in the first place, are saying to themselves, you know, we really need to have this. And when I look back, what we got four years ago, it’s not enough. So let’s go back. Let’s explain again what we’re going to do. Let’s get a new signed consent. And lots of cases, that’s fine. Sometimes they get asked to pay more money. Sometimes they have to do that. It becomes a negotiation, and that negotiation empowers the individual whose image, likeness, and voice it was. Again, that’s based on our experience. I’m not arguing that it couldn’t work out badly in some specific cases, but we do need some kind of standard, and I think that standard is one that we’ve seen as pretty effective.

 

Sonia Katyal  54:33  

Really helpful. Thank you so much, Duncan. I would love to hear Mary Anne’s take on all of this. So, Mary Anne, if I can just ask you to number one, kind of speak about your thoughts about the NO FAKES Act, but then also, if you wouldn’t mind, I’d love to hear your thoughts on two of the issues that came up in our conversation. Number one, sort of how do you view the role of consent in this process, and then number two. Do you agree or disagree, perhaps, that a notice and take down regime is satisfactory? What more would you like to see in the legislation?

 

Mary Anne Franks  55:11  

Yeah, and my answer to the first question is really tied to those two specific questions that you just brought up, and I really do want to say at this point, I have to. I really think it’s important to point out that Jennifer was the reason why I took a closer look at NO FAKES to begin with. It was a bill that I had heard about, sort of aware of. I was much more focused on the criminalization aspect that we were, you know, continually trying to agitate for in the sexually explicit deep fake category. And I can’t remember exactly when Jennifer, you and I had this conversation because you were saying there’s something you probably want to hear about this, and I think it’s such a tremendous credit to you that you kept saying it, and that at least you know the version that we have now is better than the version that was before, and that it was your perspective on this that made me think about consent primarily as a kind of you know this is going back to some what is it this is these are old school feminist debates right about is consent enough right is consent what does consent even mean right and generally what I’ve thought is consent is a floor not a ceiling right so I certainly think that things done without consent should be off the table do I ever think that consent is necessarily sufficient, and I think the answer in many cases is no, for reasons that you have pointed out. That exploitation, vulnerability, susceptibility to other interests, those things really do need to be taken into account. I think that dovetails with Duncan’s reminder to us that SAG-AFTRA isn’t just about famous celebrities with tons of money. It’s also about struggling artists and musicians who you know have less in some cases fewer resources than you know your average school teacher. So I think it’s really important to keep in mind the capability of exploitation for individuals and the kind of vulnerabilities we have to be aware of and kind of right for those. And then that the aspect that both Jennifer and Duncan have brought up, which is the public question, right? That is, what does this do to us, right? Not just to the person who’s specifically being affected, but what does it do to corrupt our ability to share reality? What does it do to destroy our ability to know what is true, to know what is false, to know what we’re looking at, right? To begin with, to affect our even our willingness to look to see whether or not something is true or false to see if we care about that right. So I think that yeah, consent is a starting point, and it’s always bad when we’re dealing with without with a lack of consent. But we should not be necessarily reassured when there’s some kind of nominal consent, and we should remember who all the parties are. It’s not just going to be the individual whose image is at the basis here, but also all the other impacts that we need to think about, and that. But the positive side of this, and I was thinking, trying to reflect on what Duncan was saying about how when you get something, and absolutely want to share that, and as someone who’s tried to get bills passed in multiple places. Yeah, you never get what you want. Really, you’re lucky if you get anything. But what you hope, I think, going back to Jennifer’s opening remarks, is that you’re not making things worse, right? So partial is good, counterproductive bad. So trying to navigate between those things is sometimes very difficult, and sometimes it’s hard to see what those counterproductive points are going to be. I don’t know if we’ll have time really to talk about the TAKE IT DOWN [Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks] Act, which is the criminalization, finally the federal criminalization of both actual authentic and intimate images, but also manipulated ones, and the Take Down provision. But we have found CCRI has found in the its own experience with this that we are not happy with the version that has gotten passed because we think it’s going to do damage.

 

Mary Anne Franks  58:43  

I think though that it’s also going to do some good because when Duncan says, “Well, we need these kinds of we need we need certain big pieces in place, right? That they can’t all if the impact of NO FAKES is that there is a general recognition by the part of the public, maybe really incomplete. Of course, it’s going to be for a 98 page bill. If the general impact is that people hesitate before they abuse someone’s image, if the general rule is that there’s a general disincentive by powerful people or gatekeepers to not act like this, that’s a win, right? That’s a huge win. So the question I always have is, what is going to be the impact of these bills on actual human behavior? Is it going to make people less likely to do the bad things and more likely to do the good things? And every time I try to read NO FAKES and I cannot keep up with these different versions, I don’t know, right? I don’t know what kind of impact that’s going to have on the average situation because we are talking about massive companies, industries, powerful agents who I don’t know what’s going to control them. I think about how when Meta was getting you know hauled before Congress and sued at multiple times and starting to lose on some of these Section 230 [of the Communications Decency Act] motions in the LA County case from a few months ago. Zuckerberg, when he’s called to testify, sails into the courtroom, surrounded by people wearing meta glasses, and has them sitting, and the judge has to tell them you have to take those glasses off and you can’t record it here, and I think this is a moment where people are suing you. They are looking to see what you’re, and this is how he acts, right? Or in the middle of being in the midst of congressional investigations, announcing that they’re going to be doing Ego4D was what they were calling at the time. This kind you can record everything around you all the time. Awesome. What is it going to take to stop those major players, right? Because once we can’t hold them back, if we can’t stop them from making these tools and putting them out there to the public isn’t really going to matter that theoretically there’s a chance for remedy on the other end. How are we going to affect people’s incentives? How do we make it painful for people to do the bad thing as opposed to incredibly profitable to do the bad thing? So I’m back to thinking about you know I don’t know what the actual answer is for NO FAKES. I tend to agree with Duncan that it’s better to have something than nothing. I’m worried that that about Jennifer’s very close critiques that sometimes some of that something could be worse than the nothing.

So I don’t really know, but I think about the last goal principle. Some of you probably know this from international human rights law. The idea that you should build your laws, build your processes, your products, your anything that you’re about to do to change right the society that you’re in, or to nudge them, or is to ask yourself how will this affect the most vulnerable person in the community? Is it going to actually help them, or is it going to hurt them? And you build from there, right? Or a version of that is in the universal design theory, right? You accommodate the person who has the most vulnerabilities, you protect everybody. But I worry that we tend to have the opposite kind of approach, which is to start with the most powerful, the most obvious, and hope that it works out well for everybody else. So I don’t know. I just don’t know if on do I think that NO FAKES will help sort of put the brakes on this culture we have that is just rocketing forward to just instrumentalizing.

 

Mary Anne Franks  1:02:01  

Everybody’s just become fodder for the machine or for the personal narrative or for the whatever it is. I don’t know, but as long as it’s going to give us a little bit more that might slow some of this down, and it’s not actually making it easier for people to accelerate the process in disturbing ways, then I’d say I’m for it. But I’m just not sure.

 

Jennifer Rothman  1:02:22  

I’d like to just chime in at that point. So first of all, I agree also with everything that Mary Anne [and] Duncan said in terms of like we’re not going to get a perfect thing. I like things to be perfect, but I recognize they can’t be. I’ve worked extremely hard to get some changes in NO FAKES, as well as some state legislation in this area. I would say that the May version of NO FAKES to me was a deal breaker and definitely worse because it constructed, it would construct a federal right that allows other people to name a living person’s digital replica, including their voice, their performance, and their likeness, and that to me should be a line in the sand. And it’s one thing if state laws are interpreted incorrectly that way; they can be challenged as violating constitutional rights. It’s a very different thing if Congress adopts such a view. And so, I think we want to be very, very careful about that. The June amendment seemed to claw that back, and you know I want to parse the many pages over again to make sure that people are adequately protected. I think that how that interplays with the problem of many other vulnerable people having their rights owned by others, or online terms of service that claim that your voice and likeness and video recordings can be reused, and how that interplays with this language is, I think, unclear and worrisome. There was also an addition in the June version that I’ve read several times over, and I’m still not quite sure what it means. Which seems to create an exception to the right for coincidental resemblance if a person doesn’t have commercial value, and so that sort of claws back what was once a very broad applies to everyone right, and so that’s a little hard, a little hard to parse. And the preemption provision, the unclear preemption provision dating back to certain laws that exist as of January, I think second, 2025, also adds to that sort of layer of uncertainty of what exactly this would do. This is not, and I haven’t taken a formal stance on this bill, except that if someone else, if it allows other people to control your identity, that should be a deal breaker. Is that I’ve heard from other people that some of the notice and take down provisions are so onerous that only very large tech companies can comply with them, and that that is a strategic business choice. And so that that raises a host of things, right? Are certain people then locking up markets because others aren’t going to be able to meet these requirements, even if they’re good actors, or you know? And then it also raises the other question, which was my sort of what can law do in this space? We have this sort of whack-a-mole problem. There are always going to be bad actors. Are they going to suddenly follow the law? Obviously, we want the big companies to be following the law, and if they’re not, they should be. Which leads me to this, my final reaction, which Mary Anne raised, which is Section 230 is sort of an elephant in the room here. Instead of like 49 to 98 page, very difficult to parse legislation that no one will understand, particularly ordinary people. Section 230. One of the main reasons that companies don’t take things down is Section 230, and there is a circuit split on whether state publicity and privacy-based appropriation claims are fall within 230’s immunity for platforms or fall within an exception for IP-based claims. The Third Circuit has said that right of publicity claims are fall in the IP exception. So Facebook, for example, can be liable for violations of right of publicity based on third-party speech. The Ninth Circuit has taken the opposite tact. One easy way that’s about a sentence, right? You could either strike Section 230, which isn’t likely to be very popular, but a one sentence conclusion that right of publicity laws and appropriation claims fall within the intellectual property exception would provide an incentive for platforms to take things down. They take things down very expeditiously with copyright, which does have the DMCA [Digital Millennium Copyright Act] takedown provision.

 

Jennifer Rothman  1:06:58  

But they also take things down pretty rapidly for trademark claims, and so without the notice and take down provision, so I think that’s another way and a shorter way to get platforms to act in this space. And it also I’m going to just harken back to something else that Mary Anne said is property the right frame, and I’m a big advocate for understanding people’s rights over their own identities as a property right, which highlights the right of exclusion. One of the reasons people don’t like that frame is because they worry about the transferability of property, but not all property is transferable, and property means something. You’re not able to take other people’s property and use it. It fits in this framework of how we treat other forms of intellectual property, and so I think it’s not an either or. It’s not commercial versus autonomy and dignity. It’s not property versus privacy or personal. It is all of the above. This is should be a profound right that each person is imbued with, and that is inalienable to them. And we can conceive of it as an exclusionary property right without diminishing its attachment to a particular person.

 

Sonia Katyal  1:08:17  

Great. Well, picking up on this sort of idea of property. I actually want to kind of shift gears a little bit, and I want us to talk. I’d love to hear your thoughts about the sort of datafication aspect of all of this, and particularly how AI and feeding of data into AI systems is going to play a role. And so, let me start with you, Duncan, if you don’t mind, I’d love to hear your thoughts about how, in your experience, how creators are thinking about the differences between, you know, what we’ve been talking about having performances recorded and sort of replicated in ways that they might disagree with, versus sort of the idea of having their performances recorded and then used for the kind of in the service of kind of training AI, having aspects of their faces, voices, or movements kind of turned into reusable data. What kinds of concerns are you seeing on the ground?

 

Duncan Crabtree-Ireland  1:09:18  

I mean, huge concerns in this area, and let me just say, not only for our members and performers. I mean, I did a whole presentation on this at the AFL-CIO [American Federation of Labor and Congress of Industrial Organizations] convention earlier this year in Minneapolis. This is a concern for all workers, not just workers who are doing, you know, work in the entertainment or creative arts type industries, because employer surveillance and the use of AI technology to essentially extract data value from the physical actions, the workplace environment that workers exist in is a huge, huge problem, and is something that I think most of the unions in the United States and probably around the world are confronting in collective bargaining, and also working on from public policy perspective, so big problem, especially because just like with the training data that’s been used to create all of the large or virtually all of the large models that are in use now, that data is just being taken. There’s no not only is there no compensation, there’s not even any form of consent, and there’s litigation, as I’m sure we all know, going on over the question of what, how does you know, how does copyright relate to that? In terms of these major players’ decision to just scrape the internet and take everything and load it up into their systems, we and I feel really strongly about this that it is not okay for, and it shouldn’t be viewed as okay as fair use or in any other way for these companies to take the intellectual property of others that was created through, you know, considerable work, skill, and effort, and then just convert it into the building blocks of their commercial business. And I always find it quite ironic how so many of these companies are incredibly protective of their own intellectual property while having zero regard for anybody else’s intellectual property in that process. So you know they’re not open sourcing their algorithms. To the contrary, they’re monetizing their algorithms and keeping them close to the vest, while they take everybody else’s material. But for our members, it is a huge concern. We have bargained over training data use of performances in every negotiation we’ve done since 2022, and it is a really, it’s a tough one. This has been one of the hardest areas to achieve anything meaningful in collective bargaining, and the main reason for that is because of the legal uncertainty in the intellectual property world. Because I do think we have been able to get the studios and streamers, video game companies, commercial advertisers, advertising agencies, and the record labels in collective bargaining to agree to a lot of limitations on what they can do with this technology, but because they themselves are uncertain that they can withhold those items from third parties who just scrape them, or that they will be prevail in their litigation against these companies, they are also reluctant to commit to protections with respect to the artists involved, because they don’t want to get sandwiched or squeezed in the middle between third parties who can do whatever they want, and then our artists saying to them, “Well, it’s not our problem; that’s your problem. So you need to be, you know, making us whole for the losses that we’ve suffered. But in the long run, this is not only a intellectual property type issue. This is also a you know a jobs issue in the sense that what you’re often asking people to do in this case is provide training data that’s designed to create things that can replace them as workers, and we don’t want to see human creativity replaced with algorithmic outputs. We’ve been super clear about that, and we have fought hard for that. And I’m really proud of the fact that in our last negotiation with the studios and streamers, just this year, we achieved provisions in which they committed to not use synthetics, i.e. generative AI created outputs to replace humans, except under very specific circumstances, and to the point of enforceability, we’ve been talking about. We also achieved an arbitral remedy for that, so that we can, in fact, there’s a whole process they have to go through if they want to do it. And if we disagree at the end of that process, we have a right to arbitrate that, and with unlimited damages as a potential remedy for that.

 

Duncan Crabtree-Ireland  1:13:38  

So I think that there are things that we can do in these areas, and I think it’s really important for us not to feel so despondent over the state of things that we just go into lockdown mode and do nothing. We often talk about how you know a certain amount of fear is productive, and then there’s a certain amount beyond that that becomes paralyzing. And I don’t want any of us to feel paralyzed by the power, the strength, the money, the influence that these tech companies and others have, because I genuinely believe while we may not be able to get 100% of what we want if we actually you know step forward, speak out, you know as Jennifer and Mary Anne have done, and a whole bunch of our members have done. If people will do that, we can impact not only what happens in public policy, but we can impact what happens with these companies. And earlier, somebody mentioned Scarlett Johansson. I mean, granted, Scarlett Johansson has a unique platform, but Scarlett Johansson, with collaboration from our union and from some other entities, was able to get voluntary a voluntary change to what was being done by OpenAI at the time. We have been able to get OpenAI when OpenAI released Sora 2. We were able to convince OpenAI to put a lot of image likeness and voice protections into Sora 2 that were not existent in the original version of Sora, and that wasn’t directly through public policy, and they aren’t a signatory to any collective bargaining agreement with us. That was through efforts and persuasion and advocacy. So I just really feel like people need to know that there you can effectuate change by persistence and speaking out. And I think the folks on this panel are a great example of that.

 

Sonia Katyal  1:15:20  

So I really appreciate your observation on that point, Duncan. I would love to hear Mary Anne’s reaction because I think Mary Anne so presciently on this panel spoke about kind of thinking about the most vulnerable and the effects of these changes on the most vulnerable. I’d love to hear your thoughts, Mary Anne, about how you would sort of incorporate the concerns about biometric data feeding into AI and how that relates to the work that you’ve been doing.

 

Mary Anne Franks  1:15:53  

Yeah, I it is the it is hard, you know. When Duncan says we don’t want that that point about fear being useful, and then you know there’s a point at which it becomes paralyzing. I think about this a lot, right? Because the more that I look at the terrible things that are happening, the more that technology races ahead, the more I think can any of this actually be stopped? And I very much want to believe that there are ways to because I have to believe it, right? It’s not to get all Samuel Beckett about it, but you know can’t go on, and I go on, right? Because you don’t have a choice; otherwise, you just you give up, right? But I do, I do keep going back to the same point, which is that we are in this place because we ignore all of the lessons, right? That history has taught us. And when I say we, it’s the people in power, because plenty of people are trying to say, hey, don’t do this. This isn’t just going to be bad for these people. It will be bad for everyone. These are not just going to, you know, whatever happens to the least among us is eventually going to happen to all of us. Although it would be really nice if people cared about the least among us just by themselves. But yes, we know that these kinds of technological harms are not going to be evenly distributed. At least not at first. It’s always going to be women and girls. It’s going to be the undocumented. It’s going to be the poor. It’s going to be you know it’s going to be racial minorities. It will be people, sexual minorities. Right? We know all of this, and that’s why I have to believe at this point, not to sound too cynical, but everyone knows this at this point. And so when a new technology is rolled out and it hurts exactly whom you think, and you have major companies saying, “Well, we never thought that our tool was going to be used for, you know, I’m shocked, shocked to find gambling going on this establishment kind of responses. You just think, you know, how long are we going to let these companies continue to break that faith and to continue to exploit human creativity and autonomy and dignity, because it feels like we’re in this game of having to claw tiny pieces of it back, and it’s just such an infuriating idea. And when we go back to the point that I think Duncan made early on about, you know, there’s this kind of truism about how we’ve got a tension between free speech and privacy, but in fact, right, these things should be seen as connected because none of us can create or express ourselves or do good in the world or just be different or be the people we want to be if we’re constantly being surveilled and exploited and having our data extracted and used in ways that we can’t control. What we are producing is a dead society essentially, right? Because if at every point we can have our resources extracted from us, then there is no there’s no guarantee that we can there is no incentive and there is no guarantee that we can be safe to speak freely or to express ourselves. And I wish that is what people understood we were losing, right? Whatever it is that this the massive machinations of the tech industry are producing. It’s mostly on the kind of, you know, on the corpses of people’s creativity and their lives and their expressions, and that is something we should be furious about. And that rage is something I wish we could harness better, because the people who are most affected by it, going back to one of your other questions, Sonia, are the least empowered to do things about it, and so we have to figure out a way to be a bigger force than this driving force that has convinced entire countries, the world, that innovation means being reckless about vulnerable people’s harms to vulnerable people.

 

Mary Anne Franks  1:19:14  

So, I don’t know what it’s going to take for that big shift to happen, but I really do hope that it’s it is happening soon. It probably happens, you know, both in small ways and then all at once. But I do wish there was more anger, and I wish there was more demands for accountability that weren’t just, oh, this one thing got really bad. But no, as an industry, as an industry, this is a rapacious, exploitative industry that has proven itself to not be trustworthy in any sense, and so it needs to be reined in, but of course we’re living in a moment where if we’re talking about being reined in by the government, well that’s just a nightmare because now we are in a techno fascist state where you’ve got the government’s interests aligned with the most the worst aspects of the tech industry, and now what do we do about that? So part of me just you know the part that. Looking for something optimistic is to say we have to figure out how we got here so that we can undo this, but that’s going to be such a big task and there’s so many different pieces of it. It does make me despair a little bit, but I’m going to try to take a cue from Duncan and just not let it be paralyzing, but to say be angrier and ask for bigger things.

 

Sonia Katyal  1:20:19  

Yeah, I think we should all be angrier and ask for bigger things. I think that’s a really lovely theme. We are coming close to time, but I do want to ask everyone on the panel one last question, which actually goes dovetails really nicely, Mary Anne, with your last observation, which is so. I maybe I’ll start with Jennifer, and then we can go to Duncan and with you, Mary Anne. But the question that I’d like for each of you to just reflect on as we close is, you know, in this spirit of ted of trying to sort of figure out different ways to address the problems that we’ve been addressing, if there’s one thing that you could maybe change about the law today to give people more meaningful control over their identities in the age of AI, I would love to hear what you think that might be and what it might look like.

 

Jennifer Rothman  1:21:19  

Thanks. I’m. I’m still. I had things to say on the last topic, so I’m still I’m still processing. So I might say one thing, and then circle back to my one thing I would change. Although they may dovetail, but I mean I think one of the hardest questions is what to do with the input of all the data, like our biometric information, Copyrighted works, and that’s a really hard question from the intellectual property perspective. It’s a really hard question from a First Amendment perspective, and what’s much easier is sort of the output questions. So I’d like us to focus on getting that part right for a variety of reasons, like one is, I just think it’s going to be very difficult to figure out where the law is going to land on the input side. The current administration is saying that, at least from the copyright perspective, none of that training is infringing. From a practical perspective, I’m not sure we can unring the bell of all these AI models that have been built on all that data, not just in the U.S. where we get to legislate, but all over the world, and so I think we may just in terms of marshaling our resources, we may want to focus on the outputs. In terms of one major thing that I would change, I might pick more than one, but the central one is assuring that people don’t lose control over their own identities, name, voice, likeness, performances, digital replicas. This is largely a matter of state law. It is uncertain. It is abused. It is weaponized. There are particularly vulnerable people who it affects the most, who don’t have unions in their corner or who aren’t able to bargain around it later in life when they become incredibly successful. This will apply to all of us. It will apply to us even more in the online context, as I mentioned, with terms of service, where we’re giving over our rights to our name, likeness, and voice, and I think there’s not a lot of incentive to protect those rights. In part, because most people are unaware of the dangers posed by losing control, or that they’re even at risk, and because people who have a lot of financial stake are trying to use every opportunity to legislate in a way that allows them to have greater control and greater commercial success without regard to the underlying individuals, and so that’s the prize. That’s where I would focus our effort is making sure that it’s not even a question that everybody, while they’re alive, gets to control their name, likeness, and voice. We can have a whole separate conversation about why we want to create billions of dollars of wealth in dead people owned by unrelated corporations. But I don’t think we have time to do that now. But I will flag that a lot of legislation creates rights only in really commercially valuable people for 70 years or more after people die, but I think we’ll stay focused on outputs and living people being able to control their own identities.

 

Sonia Katyal  1:24:32  

Thanks so much, uh, Duncan.

 

Duncan Crabtree-Ireland  1:24:35  

Oh gosh, this you know Jennifer just says so much in such a compact way. It’s amazing, and yeah, I mean, to not duplicate anything, I agree with a lot of what you just said. I would say first of all, I just want to acknowledge that there are some companies out there who are doing good things, right? So even though you hear a lot from me about companies doing bad things, you know, companies like Flawless AI, companies like YouTube developing likeness ID, for example, or even this morning I was reading a New York Times article about Anthropic blocking some use of its tools to create biological weapons. I mean, good lord, the fact that we have to even worry about this issue in the context of that is distressing. But I do think companies that do the right thing or even do good things need to be acknowledged for that because we want to create an incentive system which is not just only call them out when they do something wrong. So I wanted to start by saying that, and I think another reason for some optimism, Mary Anne, I agree with your concerns, but is the fact the public I do think is starting to recognize more the dangers of AI in general, and have more concern about this. I mean, for example, the legislation that we got enacted in New York and is now pending in California, requiring disclosure when synthetic humans are being used to market to you, has triggered a pullback from some of the companies that were eager to use synthetic humans for marketing because they don’t want to disclose to customers that they’re using synthetic humans to endorse their products because that undermines the very value of the endorsement that they’re seeking. So I think there are things we can do, and while we can’t stop it, you know, we picked up on the philosophy from early on, that we weren’t going to be able to block AI. Like nobody has the power to do that. No union has ever had the power to stop advances of technology. So what we have to do instead is channel it and direct it as best we can. And if we do that, we can mitigate a lot of the worst impacts of it, and maybe sometimes even use that technology to help us counteract some of the worst effects of it, things like likeness detection, for example. Also, I just want to say, you know, Jennifer said, not everybody has a union. No, but everybody should have a union. So, you know, if you are listening to this and you’re thinking to yourself, “I wish it wasn’t me fighting this battle by myself in my workplace, go get your colleagues together and unionize the labor movement has a higher reputation or whatever you want to call it has a higher popularity with the public now than it’s had in decades and that’s because people are recognizing that somebody has to come together to counteract the power that these billionaires and mega corporations have and you know you may like or not like unions and some of them have done bad stuff in the past, but that is a proven way to actually bring some countervailing force to this very unbalanced situation. And the last thing I want to say that we haven’t talked about yet is it is great to have panels like this where we’re talking to people who are interested in this topic and knowledgeable about it. But if you really want to create change, you have to get into rooms with the people who aren’t coming to panels like this, and so that is why at SAG-AFTRA we go to places like CES [Consumer Electronics Show]. You might say, why would SAG-AFTRA be at CES? Well, among other reasons, because when we go to CES and we do a panel there, we get to talk to a whole bunch of tech people who would never be listening to us anywhere else, or the World Economic Forum.

 

Duncan Crabtree-Ireland  1:27:59  

I have gotten the chance to be in one-on-one conversations with CEOs of major tech companies, and to the point that I can’t remember if it was Jennifer or Mary Anne made they are just so exuberant about what they can do that a lot of times they aren’t thinking about whether they should. And even a five minute conversation where you can point out some of the implications that they aren’t naturally thinking of can have a big impact, and sometimes it’s not it’s not malicious or malevolent on their part. It’s that they just don’t think about it because they are 100% focused on what it is that they’re trying to accomplish. So I just encourage all of us to take every opportunity to get into rooms that we wouldn’t normally be in and talk about these issues, so that you know perhaps all of us can help to avoid a world that ends up like the Matrix and up in a world you know where AI is doing things like curing diseases and helping us solve hunger and things of that nature. So that’s what I hope for the outcome here.

 

Sonia Katyal  1:28:57  

Thank you so much, Duncan, and for our last word, Mary Anne, what are your thoughts?

 

Mary Anne Franks  1:29:05  

Well, I get to have the benefit of what I’ve heard from Jennifer and Duncan, and I want to live in the world where those things happen because I think that’s incredible. And on that picking up where Duncan had said about the different rooms to be in, it was I went to a fascinating panel discussion last night at Georgetown that was about the Pope’s encyclical on AI, and it was we had a really interesting. It was not we, but we had a really interesting mix of people on this panel, including a tech journalist who was talking about how, in private conversations, off the record conversations with Silicon Valley tech company leaders, they would cry about the impact that they know their products are having and feeling so morally conflicted, and that they’re reading they’re reading the Pope’s encyclical that piece of writing has actually you know escaped containment, right? And people are talking about these really hard questions, and I thought, well, this is there has to be some way to harness this, right? This there has to be some way to talk about the reservations that not just the general public is having, but the people who are actually working at these companies are also having some way to encourage them to think about using technology for good and to actually do something in a big sense to ensure that we’re not just continuing to engage in the project of recklessness and maximization of profit and thinking about what you about the harms afterwards, breaking that moral hazard, right? The way that the tech industry has just come to believe that it has the right to just do whatever it wants, and if it gets caught out, if it gets criticized afterwards, maybe it will also offer some tools here, some tools there to mitigate it.

We have to break that cycle. I appreciate when companies do good, but I also want to ask: Did you create a problem and then come up with a little bit of a patch afterwards and expect to be praised for that? Because that cycle cannot continue. You should have thought about this before, and if you weren’t thinking about it, hire better people. So I think we are way past the point of giving the tech companies benefit of the doubt. Not at least none of the companies that exist today. There is no reason that they should not be thinking from the ground up about how is this going to be used for evil and planning to make sure that that doesn’t happen. I’ve suggested back in 2017 that every major technology have a kind of equality testing and public testing before it is allowed to be foisted upon the public. I know that’s unrealistic, but that I think is the kind of thing that would need to happen. You should get buy-in from the public and from experts. But the tiny thing that I would suggest, and I know it sounds outdated at this point, probably if it’s a legal fix, is actually to do something serious about Section 230. I personally believe this is a whole other conversation about how AI is not a Section 230 issue because you know Section 230 is supposed to be about user generated content, not the actual outputs of these entities themselves. But I firmly believe that that attitude, that reckless attitude that we see from the industry, that they can do whatever they want and never have to clean up it later, comes from the way that they have interpreted Section 230 and the way that it’s been interpreted. And we need to do a very serious reset of that, and say this is you are going to be held accountable for your actions.

 

Mary Anne Franks  1:32:06  

So that is one small fix that I would make to really roll back the misguided mass of immunity, the kind of blank check that Section 230 has been interpreted as providing as a way not just of letting people sue because that’s the other thing. I don’t want to hand over more power to this government, certainly. But generally speaking, it’s not really a top-down thing. I’m thinking of it’s let people bring lawsuits, let people get into the court and talk about what did these companies know, when did they know it, are they actively poisoning and hurting us, and if so, can we do something about that? 

 

Sonia Katyal  1:32:39  

This is a really, uh, uhm, Mary Anne. That sounds like a fantastic topic for next year’s symposium, and I hope that you’ll all hopefully come back and be part of our symposia going forward. This has been a really, really remarkably, just incredibly insightful conversation. I have, I feel like I’ve so much to think about in terms of the role, the role of law, the role of technology, the role of contracts, the role of unions, the role of Section 230. So I just really want to thank all the panelists for sharing their brilliant insights with all of us. And I believe that in terms, since we won’t have time for questions with the recording, if folks watching the recording do have questions, we encourage you to direct them to the individuals on this panel or to the journal, and hopefully we’ll be able to deal with them in a later way. Thank you so much, everyone. It’s been really great. Thank you.