S.Hrg.119-202

1[Senate Hearing 119-202]2[From the U.S. Government Publishing Office]34                                                        S. Hrg. 119-20256                TOO BIG TO PROSECUTE?: EXAMINING THE AI7                      INDUSTRY'S MASS INGESTION OF8                   COPYRIGHTED WORKS FOR AI TRAINING910=======================================================================1112                                HEARING1314                               before the1516               SUBCOMMITTEE ON CRIME AND COUNTERTERRORISM1718                                 OF THE1920                       COMMITTEE ON THE JUDICIARY21                          UNITED STATES SENATE2223                    ONE HUNDRED NINETEENTH CONGRESS2425                             FIRST SESSION2627                               __________2829                             JULY 16, 20253031                               __________3233                           Serial No. J-119-303435                               __________3637         Printed for the use of the Committee on the Judiciary3839                [GRAPHIC(S) NOT AVAILABLE IN TIFF FORMAT]4041                        www.judiciary.senate.gov42                            www.govinfo.gov4344              U.S. GOVERNMENT PUBLISHING OFFICE4561-891                WASHINGON : 20264647                       COMMITTEE ON THE JUDICIARY4849                  CHARLES E. GRASSLEY, Iowa, Chairman50LINDSEY O. GRAHAM, South Carolina    RICHARD J. DURBIN, Illinois,51JOHN CORNYN, Texas                       Ranking Member52MICHAEL S. LEE, Utah                 SHELDON WHITEHOUSE, Rhode Island53TED CRUZ, Texas                      AMY KLOBUCHAR, Minnesota54JOSH HAWLEY, Missouri                CHRISTOPHER A. COONS, Delaware55THOM TILLIS, North Carolina          RICHARD BLUMENTHAL, Connecticut56JOHN KENNEDY, Louisiana              MAZIE K. HIRONO, Hawaii57MARSHA BLACKBURN, Tennessee          CORY A. BOOKER, New Jersey58ERIC SCHMITT, Missouri               ALEX PADILLA, California59KATIE BOYD BRITT, Alabama            PETER WELCH, Vermont60ASHLEY MOODY, Florida                ADAM B. SCHIFF, California6162             Kolan Davis, Chief Counsel and Staff Director63         Joe Zogby, Democratic Chief Counsel and Staff Director6465               Subcommittee on Crime and Counterterrorism6667                      JOSH HAWLEY, Missouri, Chair68LINDSEY O. GRAHAM, South Carolina    RICHARD J. DURBIN, Illinois,69JOHN CORNYN, Texas                       Ranking Member70TED CRUZ, Texas                      AMY KLOBUCHAR, Minnesota71MARSHA BLACKBURN, Tennessee          CHRISTOPHER A. COONS, Delaware72KATIE BOYD BRITT, Alabama            RICHARD BLUMENTHAL, Connecticut73                                     CORY A. BOOKER, New Jersey7475               Stephen Andrews, Republican Chief Counsel76               Saurabh Sanghvi, Democratic Chief Counsel7778                            C O N T E N T S7980                              ----------8182                           OPENING STATEMENTS8384                                                                   Page8586Hawley, Hon. Josh................................................     187Durbin, Hon. Richard J...........................................     38889                               WITNESSES9091Baldacci, David..................................................     992    Prepared statement...........................................    2893    Responses to written questions...............................    8294Lee, Edward......................................................    1295    Prepared statement...........................................    3396    Responses to written questions...............................    8797Pritt, Maxwell...................................................     498    Prepared statement...........................................    5199    Responses to written questions...............................    93100Smith, Michael...................................................     6101    Prepared statement...........................................    72102Viswanathan, Bhamati.............................................     8103    Prepared statement...........................................    77104105                                APPENDIX106107Items submitted for the record...................................    99108109                  TOO BIG TO PROSECUTE?: EXAMINING THE110                    AI INDUSTRY'S MASS INGESTION OF111                   COPYRIGHTED WORKS FOR AI TRAINING112113                              ----------114115                        WEDNESDAY, JULY 16, 2025116117                      United States Senate,118        Subcommittee on Crime and Counterterrorism,119                                Committee on the Judiciary,120                                                    Washington, DC.121    The Subcommittee met, pursuant to notice, at 12:03 p.m., in122Room 226, Dirksen Senate Office Building, Hon. Josh Hawley,123Chair of the Subcommittee, presiding.124    Present: Senators Hawley [presiding], Durbin and Welch.125126            OPENING STATEMENT OF HON. JOSH HAWLEY,127           A U.S. SENATOR FROM THE STATE OF MISSOURI128129    Chair Hawley. Welcome, everyone, to the hearing today,130which is entitled ``Too Big to Prosecute?: Examining the AI131Industry's Mass Ingestion of Copyrighted Works for AI132Training.'' This is the third hearing of the Senate Judiciary133Committee's Subcommittee on Crime and Counterterrorism, which I134am delighted to work on with my colleague, Ranking Member135Durbin.136    I want to say a special thank you to the witnesses for137being here. Many of you, I think all of you, traveled in order138to be here today. Thanks to everybody for accommodating our139change in time. The Senate floor is going to be tied up here140later today, and thus, no Committee business is happening, so141thanks, all of you, for being here and for accommodating us.142    I am going to make just a few opening remarks. Senator143Durbin will do the same. Then we will swear in the witnesses144and be off to the races.145    Let me just start by saying that today's hearing is about146the largest intellectual property theft in American history.147For all of the talk about artificial intelligence and148innovation and the future that comes out of Silicon Valley,149here is the truth that nobody wants to admit. AI companies are150training their models on stolen material, period. That is just151the fact of the matter. And we are not talking about these152companies simply scouring the internet for what is publicly153available. We are talking about piracy. We are talking about154theft. For years, AI companies have stolen massive amounts of155copyrighted material from illegal online repositories.156    Now, the FBI and the Department of Homeland Security157regularly prosecute individuals who engage in exactly the same158kind of behavior using platforms like LimeWire or Napster in159the old days, using a process called torrenting. But have these160Big Tech companies been prosecuted? No, of course not. They are161getting off scot-free. And this hearing will show us that Meta162and Anthropic and other AI companies are willfully using these163illegal networks, these torrenting networks as they are called,164to steal vast swaths of copyrighted materials.165    The amount of material that we are talking about is166absolutely mind-boggling. We are talking about every book and167every academic article ever written. Let me say that again,168every book and every article ever written, billions of pages of169copyrighted works, enough to fill 22 libraries the size of the170Library of Congress. Think about that, 22 libraries of171Congresses full of works. That is how much has been stolen.172    And this theft was not some innocent mistake. They knew173exactly what they were doing. They pirated these materials174willfully. As the idea of pirating copyrighted works percolated175through Meta, to take one example, employee after employee176warned management that what they were doing was illegal. One177Meta employee told management that, and I quote now, ``This is178not trivial.'' And he shared an article asking, ``What is the179probability of getting arrested for using torrents''--illegal180downloads--``in the United States?181    Another Meta employee shared a different article saying182that downloading from illegal repositories would ``open Meta up183to legal ramifications.'' That is a nice way of saying that184what they were doing was exactly, totally, 100 percent barred185by copyright law.186    Did Meta management listen? No. They bulldozed straight187ahead. We will see evidence today that Mark Zuckerberg himself188approved the decision to use these pirated materials. And then189the best part, Meta management tried to hide it. They tried to190hide the fact that they were engaged in the illegal download of191pirated works, and not just the illegal download, but the192illegal distribution of these same works. They tried to hide it193by using non-company servers. They went so far as to train194their AI model--get this. Meta trained its AI model to lie to195users about what data it had been trained on. I mean, you talk196about an inception-level-worthy deception, training the AI197model to lie about what its own sources were. This isn't just198aggressive business tactics. This is criminal conduct.199    And I just want to point out, Meta's conduct is not an200exception. This is the rule when it comes to what is happening201right now in the AI space among these mega companies. Big Tech202operates on the model of do whatever you want and count on the203lobbyists and the lawyers to fix it later. They don't care204about the rule of law. They don't care about America. They205don't care about freedom. They certainly don't care about206working people. They care about power and they care about207money. And every time they say things like, we can't let China208beat us, let me just translate that for you. Every time they209say that, oh, we can't let China beat us, what they are really210saying is, give us truckloads of cash and let us steal211everything from you and make billions of dollars on it. That is212the translation. We are going to see that in the testimony and213the evidence today.214    Here is the bottom line. We have got to do something to215protect the people of this country. I am all for innovation,216but not at the price of illegality. I am all for innovation,217but not at the price of destroying the intellectual property of218the average man and woman in this country. We have laws for a219reason. Those laws ought to be enforced, and Big Tech should220not be above the law. Enough is enough. It is time to enforce221the law, and that is what this hearing today is about.222    Now, I will turn it over to Ranking Member Durbin.223224         OPENING STATEMENT OF HON. RICHARD J. DURBIN,225           A U.S. SENATOR FROM THE STATE OF ILLINOIS226227    Senator Durbin. Thanks, Mr. Chairman.228    The way AI interacts with intellectual property rights,229particularly copyrights, is a critical topic we can't overlook.230America's creative industries, including software, music,231movies, literature, collectively contribute over $1 trillion to232our economy each year, employing millions of people. While AI233can be an incredible tool that unlocks further creativity,234writers, artists, musicians, and others are rightfully235concerned about what technology would mean to them personally.236Should AI companies be able to use their materials freely as237``fair use'' or should they receive compensation when their238works are used to train AI models?239    I want to tell you, chapter one, how I discovered240intellectual property. I was an attorney in Springfield,241Illinois, and in a rash moment decided to buy a restaurant. So242I joined a few friends and bought a restaurant, and we had live243music. And I got a phone call one day from a fellow who said, I244just was out at your restaurant. I said, great, did you have a245good time? Couldn't have been better. Saturday night, the music246was terrific. And I said, well, I am glad you had a good time.247    And he said, you played 10 BMI tunes and six ASCAP tunes. I248said, no, I didn't, I didn't play any tunes. He said, well, the249way the law is written, you are responsible for the fact that250copyright material was used by you to make a profit at your251restaurant. I said, tell it to the judge. He said, no, before252you say that, call your friend over in Jacksonville, Illinois,253a few miles away and ask him about a similar experience. And254his reaction was the same as yours. I called my friend who255said, ask him how much money he needs each month for ASCAP and256BMI, and we started paying it. That was my first course in257intellectual property. I hold onto it to this day.258    So how can creators compete with AI products that generate259content at the push of a button, especially when the content260might mimic or even produce their own work? These are just a261few of the questions that we are going to consider in this262hearing as we try to find the right balance between promoting263technological innovation, protecting the work of our Nation's264creators, and continuing to incentivize creativity in years to265come.266    We must recognize that AI innovation and protection of267intellectual property rights are not mutually exclusive. That268is why it is troubling, as I listened carefully to the269Chairman, to hear stories about steps Big Tech companies are270taking to train their AI models on copyright materials without271compensation to the creators of these works. For example,272rather than license authors' works, companies like Meta and273Anthropic have obtained copyright materials from sites that274host pirated copies of the authors' books and writings.275Anthropic pirated over 7 million books from shadow libraries.276As Anthropic's CEO put it, Anthropic had many places from which277it could have purchased, but it preferred to steal them to278avoid ``legal practice business slug,'' whatever that means.279While Anthropic later became not so gung-ho about training280their LLM on pirated books for legal reasons, it kept the281pirated copies that it had already downloaded anyway. I don't282get that.283    As a judge in the Meta case recently put it, ``Companies284have been unable to resist the temptation to feed copyright-285protected materials into their models without getting286permission from the copyright holders or paying them for the287right to use their works for this purpose.''288    This hearing is going to be interesting. Thanks, Mr.289Chairman.290    Chair Hawley. Thank you very much to the Ranking Member.291    It is the practice of the Judiciary Committee and all of292its Subcommittees to swear in witnesses before they testify, so293could I ask you to stand up, raise your right hand, and repeat294after me.295    [Witnesses are sworn in.]296    Chair Hawley. Very good. We will now proceed to opening297statements. We will give 5 minutes to each witness. I will just298say a brief word of introduction before each witness. We will299just go straight down the table here down the dais. We will300start with Mr. Max Pritt. Mr. Pritt is a partner at Boies301Schiller, and he represents authors in a civil copyright302infringement suit against Meta, among other matters.303    Mr. Pritt, the floor is yours.304305  STATEMENT OF MAXWELL PRITT, PARTNER, BOIES SCHILLER FLEXNER306                 LLP, SAN FRANCISCO, CALIFORNIA307308    Mr. Pritt. Chairman Hawley, Ranking Member Durbin, thank309you for the invitation and opportunity to testify today. The310Art of the Deal by Donald Trump, Hillbilly Elegy by J.D. Vance,311Theodore Roosevelt: Preacher of Righteousness by Josh Hawley,312these are just a handful of the many, many millions of313copyrighted books and publications that some of the world's314largest and wealthiest corporations--Meta, OpenAI, Anthropic,315and others--knowingly and intentionally pirated from illicit316online marketplaces for financial gain and to seek a317competitive advantage in AI.318    Today, this Committee begins to investigate and shine a319light on what is likely the largest infringement of American320intellectual property by U.S. companies in our Nation's321history. As tech companies scrambled to release generative AI322models and to catch up with OpenAI's ChatGPT, many of them323turned to illicit online repositories to take tens of millions,324if not hundreds of millions, of books and scholarly325publications and articles for free instead of buying them or326licensing them from copyright owners. By pirating these works,327AI companies have built a multibillion-dollar industry that is328projected to be a trillion-dollar industry in the next few329years without paying a single cent to the authors whose works330power their products or the publishers responsible for331introducing those works to the public here and abroad.332    Take Meta, for example. From the early days of its333generative AI program, Meta concluded that training its models334using books and articles would help their performance. But335instead of buying or licensing these works from copyright336owners, Meta decided to take them from notorious online337marketplaces of stolen copyrighted works, including some of the338same ones targeted by the Department of Justice and the FBI for339criminal copyright infringement. And Meta didn't just download340books from these illegal repositories. It used the same kind of341peer-to-peer file-sharing networks that powered Napster. In342other words, Meta also made copies and sent them to other343pirates.344    In total, Meta pirated well over 200 terabytes, terabytes345of pirated books and articles, a size comparable to the entire346printed collection of the Library of Congress 20 times over, or347the equivalent of a stack of many billions of pages of text.348Meta's piracy included many millions of works, including at349least 12 books authored by Members of this very Subcommittee350and every U.S. President and Vice President in the 21st351century. Meta also made and sent copies of over 40 terabytes of352pirated works to others.353    In doing so, Meta has helped to revive online piracy by354propping up the foreign criminal syndicates that run these355illicit marketplaces to violate U.S. copyrights around the356globe. As Anna's Archive, the largest illicit online357marketplace of stolen literature in the world today, says on358its own website, ``Shadow libraries were dying. Then came AI.''359    Meta is not alone, and it was not the first U.S. company to360engage in rampant domestic piracy for its own commercial361purposes. Pending lawsuits against OpenAI and Anthropic362revealed that both companies also pirated millions of363copyrighted works. And the decisions to engage in this mass364domestic piracy were made at the highest levels. Company365documents that are now public show, for example, the decision366to pirate instead of license was approved by Meta's co-founder367and CEO, Mark Zuckerberg, himself.368    This decision to engage in mass piracy was made, even369though key employees knew that doing so was both illegal and370unethical. One Meta researcher argued that using pirated371material should be beyond our ethical threshold. Another called372Meta an accomplice to piracy. Yet another warned that if the373media got wind of the company's use of pirated data, it could374undermine Meta's negotiating position with regulators, the very375people in this room and across the hall, in the White House,376and in State houses across the country. And when asked if he377cared whether Meta protects human creativity rather than378exploits it, Meta's head of AI partnerships testified, he does379not care.380    AI companies now seek a pass for this unprecedented piracy381by invoking a limited exception to copyright infringement382called fair use, which Congress codified in the Copyright Act383of 1976. They also argue they can't compete with China if they384can't infringe every American's copyright. Nonsense. Our tech385companies employ the best and brightest minds in the world, and386they are the wealthiest corporations in the world. It is not387credible for these companies to argue they can invest hundreds388of billions of dollars into hiring talent and building data389centers to power their commercial AI products and models, but390they can't pay a single cent to copyright owners. There is no391carveout in the Copyright Act for AI companies to engage in392mass digital piracy.393    I am grateful to Chairman Hawley, Ranking Member Durbin,394and this Subcommittee for your attention to the issue. I look395forward to your questions.396    [The prepared statement of Mr. Pritt appears as a397submission for the record.]398    Chair Hawley. Thank you very much.399    Next up is Professor Mike Smith. Professor Smith is400professor of information technology and marketing at Carnegie401Mellon University. He has written extensively on piracy and its402effects on innovation. Professor Smith.403404STATEMENT OF MICHAEL SMITH, PROFESSOR OF INFORMATION TECHNOLOGY405    AND MARKETING, CARNEGIE MELLON UNIVERSITY, PITTSBURGH,406                          PENNSYLVANIA407408    Professor Smith. Chairman Hawley, Ranking Member Durbin, I409am very honored and thankful for the opportunity to testify410today on this important issue. My testimony today is informed411by 25 years of empirical research into the impact of new412technologies on the markets--on the creative markets and my413experience serving on a roundtable of 10 economists convened by414the U.S. Copyright Office to study the implications of415generative AI on copyright policy.416    My research into piracy started in the early 2000's when417digital piracy was a relatively new problem for the creative418industries. During that period, many in the tech community419argued that piracy was fair use because it would not harm legal420sales, was unlikely to harm creativity, and any legislative421efforts to curtail piracy would not only be ineffective, but422would also stifle innovation.423    My empirical research over the past 25 years has studied424these questions. In 2020, my colleagues and I surveyed over 40425papers published in peer-reviewed academic journals as part of426a piracy landscape study we wrote for the U.S. Patent and427Trademark Office. Our report drew three broad conclusions.428    First, the peer-reviewed academic literature shows that429digital piracy does harm creators by reducing their ability to430make money from their creative efforts.431    Second, the peer-reviewed academic literature shows that432digital piracy does harm society by reducing the economic433incentives for investment in creative output.434    Third, the peer-reviewed academic literature shows that435copyright enforcement has been effective in reversing these436harms while also allowing businesses and legal online437distribution platforms to thrive and innovate.438    Today, we're hearing many of the same arguments we heard in439the early days of the internet. Allowing generative AI440companies to use pirated content to train their models is fair441use because it won't harm legal sales, won't harm creativity,442and any enforcement efforts to curtail the use of pirated443material for training will not only be ineffective, but will444also stifle innovation.445    My response to those arguments is that while the time has446changed, the underlying economic principles are the same today447as they were in 2000. And by applying those principles, I think448we can draw many of the same conclusions.449    First, the use of pirated content to train generative AI450models will harm sales for creators. Allowing generative AI451companies to train their models with pirated content is likely452to harm markets for creators by damaging the original markets453for their work, by damaging licensing markets for those works,454and by creating perverse incentives for bad actors to add new455copyrighted content to pirate networks, in essence, allowing456generative AI companies to launder licensable content through457piracy.458    Second, the use of pirated content to train generative AI459models will harm society by reducing economic incentives for460creators. This conclusion is similar to the early piracy461research: Economic incentives drive creative output. But there462is a new and unique indignity to our current situation. When463piracy is used to train generative AI models, we're not only464stealing from creators, we're then using the theft of their465content to create tools that can flood the market with machine-466generated output, which in turn will replace many of those467creators, particularly emerging artists.468    And third, as in the early days of piracy, I believe that469enforcing copyright law in the context of generative AI470training can be effective at reversing these harms and can471create a world where both the creative industries and the472technology industries are able to thrive. If the Napster and473Grokster decisions had gone the other way in the early 2000's,474it is hard to imagine that Spotify and Netflix would exist475today, and that would be to the detriment of consumers, the476creative community, and the technology community.477    I think today we have a similar opportunity to create a478win-win-win for society, creators, and tech firms by making it479clear that piracy is wrong and that a vibrant technology480economy depends on a vibrant creative economy. We found a way481to make licensed streaming and sales channels work for482consumers, copyright owners, and platforms in the early 2000's.483We must do the same for generative AI today.484    Generative AI has the potential to benefit industry and485society in many ways, but achieving that potential will require486a more robust and transparent partnership between technology487firms and the creative industries. On our current path, we risk488killing the goose--or in this case, the authors, musicians,489coders, and filmmakers--who laid the golden eggs that are key490to the present and future value of generative AI output.491    I thank you and look forward to your questions.492    [The prepared statement of Professor Smith appears as a493submission for the record.]494    Chair Hawley. Thank you very much.495    Next up is Professor Bhamati Viswanathan. Did I get that496right, Professor? Am I close?497    Professor Viswanathan. Perfect. Thank you.498    Chair Hawley. Okay.499    Professor Viswanathan. Perfect.500    Chair Hawley. Professor Viswanathan is a professor of law501at New England Law School, and she is an expert in AI and502copyright. Thank you for being here. The floor is yours,503Professor.504505STATEMENT OF BHAMATI VISWANATHAN, PROFESSOR OF LAW, NEW ENGLAND506               LAW SCHOOL, BOSTON, MASSACHUSETTS507508    Professor Viswanathan. Chairman Hawley, senior Ranking509Member Durbin, and Members of this Subcommittee, thank you so510much. I am honored to testify today on a subject that I feel511passionate about.512    I feel that Senator Hawley did an excellent job of laying513the table for us. I would like to drill down on what he's514presented us with so far and help us walk through this.515    So first, it's an interesting moment that we're at.516Generative AI is a promising set of technologies, and I think517we can all agree that they're beneficial. However, the training518that they're engaging in is deeply problematic and troubling,519and courts don't know what to do about this yet. They haven't520reached a consensus on what should obviously be done about the521training of AI on pirated works. So I would like to give us a522call for action and a solution as I talk us through this.523    First, we know that what pirate websites are doing is524illegal. How do we know that? Multiple actions have been525brought against pirate sites, and in every case, the pirate526websites or repositories have lost. The FBI, the Department of527Homeland Security have gone after pirate websites and tried to528shut them down. Now, of course, we all know this can be like529whack-a-mole, right? They shut down, they come back up again.530But the point is it's well-established that what they're doing531is illegal, and that makes sense.532    If you and I stole books from the library or from a533bookstore and said, I need to train, I need to learn, I need to534develop my mind, we wouldn't argue that this is fair use. We'd535say you can't steal the materials even for a good cause. That's536not even what's happening here. The AI generative companies are537going to pirate websites, stealing the materials that have538already been stolen. It is a crime compounding a crime. How is539this fair?540    Say you want to go drag racing, an illegal activity. I tell541you, hey, there's a shop down the street that sells stolen cars542for cheap. Go buy a car and you can drag race. You go, great,543that helps me be able to afford what I want to do. You buy a544stolen car, you drag race, you win. Do you now get to say, hey,545it's okay that I stole that--I bought that stolen car. It's546okay that I engaged in illegal activity. Neither activity is547legal, and one is compounding the other, and that's what's548happening here. It's simple. It's a crime compounding a crime.549    And it's not a victimless crime. As Professor Smith showed550us, there are real victims here, the loss of author's551livelihoods. Mr. Baldacci will be eloquent on this topic, but552as an author myself, I feel the same. The loss of my livelihood553not only hurts, but it affects what I have spent my life554training to do.555    It contravenes copyright laws, basic incentive structure. I556don't just teach copyright, I teach constitutional law as well.557This is enshrined in the United States Constitution. The558Intellectual Property Clause is one of the things that makes559this country not just great, but robust, powerful, economically560hugely successful. Over $1 trillion in revenues from the561creative content industries, this is truly at risk right now,562this entire incentive structure that was brilliantly thought of563by our Founders.564    It has negative incentives. If you know that you can go to565a pirate website and steal things, why would you ever pay for566anything again? The generative AI companies have shown us the567way to massive theft, not just by themselves, but by others as568well. It depreciates the quality and the quantity of works out569there. The tradeoff of copyright law is you, the copyright570author, take the risk, and the market rewards you with rewards571if the marketplace likes what you've done. There is no572incentive structure anymore. That's been undermined by what's573happening now.574    And there's a solution. The solution is licensing. It575already exists, the licensing of works, the fair compensation576of creators. These are all things that actually exist now. We577don't even need new legislation in some ways. We might want578that as well someday, but right now we have a solution. Enforce579good, standard, accepted, acknowledged licensing practices.580    None of this is to say that we're against innovation. We581all believe in innovation. We believe that generative AI has582potential. But you cannot compromise the livelihood of583creators. You cannot compromise our trove of creative activity584and our entire world of art and culture and the things that we585have done that make us most human and that enrich us the most--586you cannot compromise those simply by saying we need new587technologies to flourish. What we need is for new technologies588to flourish fairly, sustainably, in ways that make sense to us589and that have already been provided for by our Constitution, by590the U.S. copyright law, by intellectual property law itself.591    It is critical that Congress recognize that this is the592tradeoff that matters for the livelihoods of everyone whose593lives right now and well-being are at risk.594    Thank you so much.595    [The prepared statement of Professor Viswanathan appears as596a submission for the record.]597    Chair Hawley. Thank you very much, Professor.598    Next is Mr. David Baldacci. Mr. Baldacci is one of the599best-selling authors in America. I don't know how many books he600has had as the number one New York Times bestseller. I bet he601knows. Maybe he will tell us. I have read his books. I am602delighted to have him here today. He is going to tell us about603AI's impact on authors. Welcome, Mr. Baldacci.604605                 STATEMENT OF DAVID BALDACCI,606             BESTSELLING AUTHOR, RICHMOND, VIRGINIA607608    Mr. Baldacci. Thank you. It's a lot, number one, best-609selling.610    [Laughter.]611    Mr. Baldacci. I'll leave it at that.612    Chairman Hawley, Ranking Member Durbin, Members of the613Subcommittee, 119 years ago, Mark Twain traveled to D.C. and614appeared before a Congressional Committee to advocate on behalf615of copyright--stronger copyright laws. He was the most pirated616author of his day. I'm pirated all over the world as well. I617get why that upset him. He thought creative arts was the618lifeblood of this country, and I agree with him. That was the619first time at that hearing that he wore his signature white620suit publicly, and he did so because he thought it represented621purity of thought and spirit. I don't own a white suit, and622even if I did, I don't think my wife would have let me wear it623today, so you just get blue.624    Twain once said that ``Travel is fatal to prejudice,''625meaning if you meet people where they live, you find out626they're just like you. I had no chance to leave the segregated627world of Richmond, Virginia, when I was growing up, but I628visited the library every week, and I liked to think through629books. I traveled the world without a plane ticket or a630passport. And born from my love of reading came my desire to be631a writer.632    I worked away for decades and getting rejected over and633over, but I kept going, honing my craft, remaining disciplined,634taking the rejections head on, and using them as motivation,635and finally I was successful. And after 60 novels under my636belt, I work just as hard as I ever have. It's the American637way. You work hard, you play fair, you stay the course, and638you'll make it.639    I truly believed that until my son asked ChatGPT to write a640plot that read like a David Baldacci novel. In about 5 seconds,6413 pages came out that had elements of pretty much every book642I'd ever written, including plot lines, twists, character643names, narrative, the works. That's when I found out that the644AI community had taken most of my novels without permission and645fed them into their machine learning system. I truly felt like646someone had backed up a truck to my imagination and stolen647everything I'd ever created.648    I'm aware of the argument that what AI did to me and other649writers is no different than an aspiring writer reading other650books and learning how to use them in original ways. I can tell651you from personal experience that is flatly wrong.652    I was once such an aspiring writer. My favorite novelist in653college was John Irving. I read everything that Irving wrote.654None of my novels read remotely like a John Irving novel. Why?655Well, unlike AI, I can't remember every line that Irving wrote,656every detail about his characters and his plots. The fact is,657also unlike AI, I read other writers not to copy them or steal658from them but because I love their stories. I appreciate their659talent. It's motivated me to up my game.660    What AI does is take what writers produce as an incredibly661valuable shortcut. It's like super fuel to teach software662programs what they need to know. And I have learned that these663trillion-dollar companies didn't even buy my books. They got664them off a website that has pirated works. They complained that665it would be far too difficult to license the works from666individual creators, so apparently, it was more efficient to667steal it. Trillion-dollar companies with battalions of lawyers668did not have the resources to do things lawfully.669    I was once a trial lawyer. If I had made that argument in670court, I would either have been laughed out of the courtroom or671held in contempt by the judge and rightly so. If AI companies672only needed words, they could have fed every dictionary in the673world into their machine learning, but that was not nearly good674enough because it would mean decades of additional work and675hundreds of billions of dollars of additional investment. What676they needed was complete, well-crafted, living, breathing677stories with characters that seemed real, plots that made678sense, dialog that appeared genuine, humanity on the page. In679sum, they needed us and our craft that we learned with the680sweat of our brows and the flexing of our imaginations.681    And these companies have swooped in, stolen that labor in682order to make enormous profits. But we, the writers, the true683source of all of this, will receive nothing. AI will allow684anyone, with no effort at all, to order up a novel written in685the vein of an established writer. And that book can be sold686saying that it reads just like a David Baldacci novel. Yes, it687does read like my novels because it is my novel. It is my688imagination.689    People complain about cheap imported goods hurting American690workers. Well, we have cheap books being created by American691technology flooding the market. That will mean lower profits692for publishers and less money to spend on new emerging writers.693Trust me, that hurts all of us.694    Online vendors now require the author to disclose if a book695was not human-created. It's getting to the point where they696will have to limit the number of books that someone can publish697on a weekly or even daily basis. This is insane.698    Source code and elements of algorithms are also protected699by copyright. I would hazard to bet that if I stole any of the700AI community's source codes or algorithms and then tried to701profit off them, they would unleash a tsunami of lawsuits702against me. However, if, as AI contends, fair use is actually703my entire body of work, there is no more copyright protection704for anyone. I'm sure AI believes that their IP should be fully705protected against interlopers, and I agree with them. Thus, I706am deeply disappointed they don't feel the same about people707like me.708    The AI community apparently is there entitled to steal our709work product despite it being copyrighted because what they're710doing is so transformational. Well, let me tell you, billions711of people have been transformed by books. Many significant712events in human history and in this country had seminal authors713in their works that wrote at the head of the pack. We didn't714truly emerge from the dark ages until the invention of the715printing press when books became widely available. Books also716teach empathy, making the world a kinder, gentler, more717meaningful place.718    I'm only one man, but books transformed my life, propelling719me to a far better existence. I am sure there are aspects of AI720that will also transform the world, but if you want to bet on721which side is more transformational for all of us, I will bet722on books every single time.723    Thank you.724    [The prepared statement of Mr. Baldacci appears as a725submission for the record.]726    Chair Hawley. Thank you very much, Mr. Baldacci, very well727said.728    Next up and finally is Professor Edward Lee. Professor Lee729is professor of law at Santa Clara University School of Law,730where he has written extensively about the intersection of AI731and copyright law.732    Thank you for being here, Professor Lee.733734    STATEMENT OF EDWARD LEE, PROFESSOR OF LAW, SANTA CLARA735       UNIVERSITY SCHOOL OF LAW, SANTA CLARA, CALIFORNIA736737    Professor Lee. Chair Hawley, Ranking Member Durbin, and738other Members of the Subcommittee, thank you for this739opportunity to testify. I am a professor of law at Santa Clara740University School of Law. I'm also a book author and a741photographer, and my personal experience informs my scholarship742and understanding of the importance of copyright to authors and743artists across the country.744    In my testimony, I will discuss whether using copyrighted745works to train AI models is a fair use, giving particular746attention to the two recent decisions by Judges Alsup and747Chhabria in cases filed by book authors against Anthropic and748Meta. This novel question of law, which has important749implications for U.S. national interest, has sparked sharp750disagreements among parties, stakeholders, and now Federal751judges. As Judge Bibas noted in an earlier non-generative AI752case, this question of law is difficult.753    In my opening remarks, I would like to stress three points.754First, I believe Judges Alsup and Chhabria correctly concluded755that the use of copies to--the use of copies of works to train756an AI model serves a highly transformative purpose in757developing a new technology under factor one of fair use.758During training, an AI model is exposed to vast training759materials, typically many millions of works. Through a process760called deep learning, the model identifies the statistical761relationships among words and within subparts of words, thereby762enabling the model to conduct numerous functions, including763research, translation, delivery of medical advice, generation764of content, and so forth.765    As Judge Chhabria concluded in his opinion, ``The purpose766of Meta's copying was to train its large language models, which767are innovative tools that can be used to generate diverse texts768and perform a wide range of functions.'' And as Judge Alsup769recognized, ``The technology at issue was among the most770transformative many of us will see in our lifetimes.''771    Now, the history of AI development strongly supports this772conclusion. It is important to understand why AI researchers at773universities began training AI models on large datasets. This774practice originated not at AI companies, not at Big Tech, but775at universities where AI researchers discovered a key insight.776Scaling or using larger and more diverse datasets actually777worked in developing and improving AI models, an achievement778that escaped researchers for many years. This seminal779breakthrough, which took decades to figure out, has propelled780the advances of AI that we are witnessing today.781    Second, while I agree with the ultimate findings of fair782use in both cases, it's important to remember that fair use is783fact-specific and decided on a case-by-case basis. In some784cases, a transformative purpose in AI training might be785outweighed by the other factors. For example, an AI model that786routinely produces outputs that are infringing, such as787regurgitations, might not be a fair use even in the training of788the model due to insufficient guardrails on the model.789    Critically, in the cases against Anthropic and Meta, the790judges concluded the plaintiffs did not show the models had791produced any infringing outputs of the plaintiff's works. And792that can be appealed, but that is the findings of both judges.793    My final point is the need for caution, caution by the794courts, caution by Congress, and the States. I believe it's795important to weigh the United States' interest in AI796innovation. President Trump has issued an executive order797making U.S. development and global leadership in AI a national798priority. China has its own priority and a plan of surpassing799the United States and becoming the world leader in AI by 2030.800The United States' national priority in AI counsels caution.801    Indeed, in Google v. Oracle, another technology fair use802case of national importance, the U.S. Supreme Court itself803cautioned, ``Given the rapidly changing technological,804economic, and business-related circumstances, we believe we805should not answer more than is necessary to resolve the806parties' dispute.'' Judges Alsup and Chhabria departed from807this approach in some controversial parts of their opinions808that were just dicta. I disagree with Judge Alsup's suggestion809on pirated books and Judge Chhabria's suggestion on copyright810dilution, as more fully elaborated in my written statement.811    At this juncture, I think the best approach is for Congress812to wait and see how other district courts, the courts of813appeals, and potentially the U.S. Supreme Court resolves these814difficult issues, including access to pirated shadow libraries815in the many pending copyright lawsuits across the country.816    Thank you, Senator.817    [The prepared statement of Professor Lee appears as a818submission for the record.]819    Chair Hawley. Thank you very much, Professor. Thanks for820being here. Thanks again to all of our witnesses.821    We are going to now have 7-minute rounds of questioning,822and we will see if we can fit in maybe a couple of rounds, just823depending on the time that we have. I will start, and then we824will go to the Ranking Member and any other Members who arrive825in that time.826    Professor Viswanathan, let me just start with you, if I827could, and let's see if we can just drill down on some of the828specifics here. Mr. Baldacci mentioned in his opening statement829that AI could just feed dictionaries into their platforms in830order to train them. They don't do that. They prefer published831works, fully formed works. Why is that? Can you give us an832insight into that?833    Professor Viswanathan. That's absolutely right. They learn834syntax, structure. They learn how we learn language, right?835When you learn language, you just don't learn words. You don't836memorize words. You don't memorize notes when you learn music.837You learn structure and syntax. And the point that Professor838Lee is making is correct. They need large datasets. More is839better to learn predictive language models. However, more is840not everything. It's not pirated works.841    Chair Hawley. So let me just ask this. You said that they842are not buying the books. They are not buying Mr. Baldacci's843book or anybody's book who is sitting up here, anybody in the844audience. They are getting them. They are stealing them. They845are pirating them from somewhere. If they are not buying the846books, they are not stealing them out of libraries, where are847they getting them?848    Professor Viswanathan. These large repositories of849materials that are available online, there are many. Some are850licit, some are not licit. The pirate websites in particular851are not licit. So if you need a lot of material, you go out and852you scoop up all that material that you can find, but you don't853go to pirate websites to get that material if what you want to854do is legal. None of these works are licensed. None of these855works are licensed. No author has been compensated to date.856    Chair Hawley. They go to these--let's call them shadow857libraries--to get the works illegally. By the time they go to858the shadow library, the works there are already stolen, right?859They have already stolen Mr. Baldacci's book, Professor Lee's860book, everybody's, your books. They have stolen them. When they861go to the shadow library, how do they get them? I mean, how862does the AI company then take possession of the particular863work?864    Professor Viswanathan. There's a process called torrenting,865and I will not trouble you all with the details of torrenting,866but essentially huge amounts of data streamed to you and you867get them. At the same time, you can send them out. That's868called seeding. You can send them out at the same time.869Uploading and downloading exists at the same time. This is a870peer-to-peer process. So not only are you taking in these871pirated materials, you are also distributing them. The872violation of copyright law exists at the reproduction of these873works, at the making available of them by the pirate libraries,874the dissemination of them, and your dissemination gen AI875company of them as well.876    Chair Hawley. So they are both taking the works and877distributing them as well in this thing called, kind of like878Napster, this thing that you call torrenting. Let me ask you879this. I mean, is torrenting legal? That is not legal, is it?880    Professor Viswanathan. Torrenting can be illegal, but in881this case, it is not. And in this particular case, this is882benefiting the--now I agree with Judge Alsup who said, if883you're taking it from pirate libraries, no way. That is not884acceptable, right? Part of what we're seeing here, Judge885Chhabria said, well, it's not helping the pirate websites.886Well, yes, it is. The pirate websites, there's one in887particular called Anna's Archive. They actually put on their888website, hey, gen AI companies, come train on us. We'll do some889data swaps. Or, you know what, you can make us a donation too.890This is directly helping the pirate websites thrive, flourish,891proliferate.892    Chair Hawley. Let me ask you this. Have there been, to your893knowledge, any criminal enforcements against these torrenting894platforms?895    Professor Viswanathan. Yes, there have been attempts to.896Again, it's like a game of whack-a-mole. You get one, you knock897it down, it pops up again in some jurisdiction that you don't898have control over.899    Chair Hawley. What is the key to criminal enforcement? You900know, civil versus criminal in this context, when do we have a901criminal case against torrenting? What is the key to that?902    Professor Viswanathan. Okay. This is a really important903point. What's criminal here? Criminal copyright liability has904two prongs to it. Prong one is you have to do it willfully, and905prong two is you have to do it for commercial advantage or906gain. We clearly know that prong two is met. This is for907commercial advantage or gain. I don't think Meta is doing this908out of the goodness of its heart. Prong one, willful means you909need to know that what you are doing is illegal. There's lots910and lots of evidence now, particularly from the Kadrey v. Meta911case, that shows that they knew this was illegal. They even had912to ask all the way up the chain of command to Mark Zuckerberg913and say, hey, is this okay? And he said, yes, it's okay.914    So not only did he do it knowing it was illegal, he did it915knowingly, he did it willfully, intentionally, and whether or916not he knew what statute it was legal doesn't matter. For this917to be willful, you have to know that what you're doing is918wrong, and this meets that prong. So this is, in fact,919amounting to what you might call criminal copyright liability.920    Chair Hawley. Mr. Pritt, let me just ask you about this,921about the willful aspect, and let's talk about Meta in922particular, since Professor Viswanathan just mentioned Meta.923They are one of the biggest monopolists in the world and one of924the biggest AI companies now in the world, if not the biggest.925So let's just talk about them for a second. Meta uses torrents926to acquire pirated data for its Llama model, is that right?927    Mr. Pritt. Correct.928    Chair Hawley. How much data would you estimate that Meta929has torrented? It is illegally downloaded and also then shared930in this peer-to-peer scheme.931    Mr. Pritt. It has pirated well over 200 terabytes of932copyrighted material from multiple--I don't call them shadow933libraries because they're not libraries--but illicit criminal934enterprises.935    Chair Hawley. And how much has it paid the copyright936holders for these works that it has used, to your knowledge?937    Mr. Pritt. Nothing.938    Chair Hawley. Nothing, zero. So billions of works, billions939of books like Mr. Baldacci's, zero payment. If Meta were to940pay, do you have any idea what the cost might be? I mean, to941your knowledge and your discovery, did they ever explore942paying? I mean, is there any sense of how much this might have943cost them?944    Mr. Pritt. Early on, they explored licensing. They assigned945two individuals part-time to attempt to license, and they946decided it would take too long, for example, and that's when947they turned to piracy. At the time, they had public documents948show that certainly tens of millions, if not hundreds of949millions, had been contemplated for licensing at that time.950    Chair Hawley. Okay. So let's just think about this.951Hundreds of millions of dollars, that is the value, maybe sort952of the base, the bare value of the works that they have used,953like the works that you all have written on this panel,954hundreds of millions, and they paid zero of that.955    So let's just drill down a little further. Did Meta know956what they were doing was wrong? Do you, Mr. Pritt, believe in957the evidence you have seen that there is any evidence to958suggest that Meta's employees knew what they were doing is959illegal?960    Mr. Pritt. I think the documents that have become public961clearly show that.962    Chair Hawley. Let's just look at a few of these documents.963I am going to show you a few things, and I will ask you to help964me interpret them to make sure that we get them right. Let's965start here with a Meta employee, a Meta engineer working on966their AI project, Eleonora Presani. She says, ``I don't think967we should use pirated material.'' This is in a chat with other968Meta employees. ``I don't think we should use pirated material.969I really need to draw a line there.'' She goes on, ``I feel970that using pirated material should be beyond our ethical971threshold. Sci-Hub, ResearchGate, LibGen are basically like972Pirate Bay or something like that. They are distributing973content that is protected by copyright, and they are infringing974it.'' How do you read this, Mr. Pritt? Does this look like975knowledge to you?976    [Poster is displayed.]977    Mr. Pritt. That's certainly what we've argued in the case.978    Chair Hawley. Let's look at another Meta employee. Here is979Nisha Deo in the same chat. She replies and said, ``It's the980piracy (and us knowing and being accomplices) that's the981issue.'' This is a Meta engineer working on their AI project.982``It's the piracy (and us knowing and being accomplices) that's983the issue.''984    [Poster is displayed.]985    Let's look at another one. Here is the response that986another Meta engineer in the same chat gave. ``Well, we want to987buy books and be nice, open people here. But, however, to make988it happen and not letting the bad guys win''--that's the beat-989China argument--``we need to make a case--fast--and cut some990corners here and there.'' ``We need to cut some corners here991and there.'' Mr. Pritt, what are we looking at here? I mean, is992this knowledge of illegal activity?993    [Poster is displayed.]994    Mr. Pritt. When they refer to bad guys, I think they're995actually referring to OpenAI and other AI competitors.996    [Laughter.]997    Mr. Pritt. But yes, this is certainly one of the many998documents that show that they knew these were pirated websites999that contained copyrighted materials, and they were taking them1000for free.1001    Chair Hawley. So here we have it in black and white. Don't1002believe me. Read the evidence. These are Meta's own engineers,1003Meta's own employees saying, they know what they are doing is1004ethically wrong, illegal, likely to subject them to legal1005liability, and they are doing it anyway because they need the1006money.1007    There is a lot more here. We will come back to this. I want1008to give Senator Durbin a chance to ask questions. Senator1009Durbin.1010    Senator Durbin. Thanks, Mr. Chairman.1011    I want to ask startup questions with Mr. Baldacci. A number1012of authors have shared with the public the process they go1013through to write a book. I believe John Irving in The Imaginary1014Girlfriend did that. I think John McPhee has done that in the1015past. Stephen King has done that. Give us a kind of an insight,1016now that you have published successfully in volume, what the1017process is in writing a novel.1018    Mr. Baldacci. Well, you know, one, you have to sort of be1019in love with words and storytelling because that is sort of the1020essence of what you're trying to create. You draw upon personal1021experiences, your own curiosities, people you've met along the1022way, things that have happened to you, places you've traveled1023to, humanistic experiences that a software platform really1024can't replicate. And if it ever manages to do it, I would like1025another planet to live on, quite frankly.1026    And for me, it was 20 years of hard work learning the craft1027before I ever was published at all. I started writing short1028stories and wrote them for 15 years when I was in college and1029law school and tried to get them published and was not1030successful. But it's a craft that you build over time. And you1031have a lot of frustration, a lot of dips and valleys. Good1032times happen, bad times happen, rejections happen. You learn1033from them, you keep going. And at the end of the day,1034hopefully, you get good enough to where someone who has the1035ability to make your career happen will read your material and1036respond to it, and you can then maybe hopefully write for a1037living. And that's what happened to me after a long period of1038incubation.1039    You never really see a lot of young writers--you know,1040you're not going to see a lot of teenage writers making it big1041because writing is about life, and you have to have something1042to sort of write about. And it takes a long time. And that is1043why I felt when my son brought this up where every single one1044of my books was presented to me in an outline in like 31045seconds, it really felt like I had been robbed of everything my1046entire adult life that I had worked on now was in the1047possession of someone else that someone else I didn't even know1048could then use to write their own books that are actually my1049books. I mean, that's not supposed to happen in this country.1050    And that's what was so enraging to me that I--I license my1051work all over the world. I license it for different foreign1052publishers. I license my work for television and movies and all1053types of endeavors. And I am open to any offer. If someone1054comes to me and wants to license my work, I will listen to1055them. If we can negotiate something that's agreeable to both1056parties, I will do it, and they can use my work for the1057parameters that are in the licensing agreement, and life can go1058on and people can be happy.1059    But the uncertainty of like stealing stuff from pirated1060sites operated in Russia just so you can gain an advantage and1061you don't really care about what happens to the likes of me and1062other writers coming up--I make a lot of money from my1063publisher, and my publisher has used that money to take risks1064on new writers coming up they ordinarily would not have been1065able to take a risk on. So when you hurt established writers1066like me, you hurt all the other writers coming behind us.1067    Senator Durbin. So when you are in the creative process of1068writing novels and other things, are you policing against1069plagiarism?1070    Mr. Baldacci. I get--I am pirated a lot, but I never worry1071about that because my ideas are my ideas. And I--nobody has the1072sort of mindset and the experiences that I have, nor do I have1073the mindset and experiences of other people. It is very1074individualized. I never worry about that I'm going to1075inadvertently take something away from another writer because1076my stories are my own.1077    And that's why a software platform, the only thing they can1078do is take from what has already been created. They can't1079create anything really on their own. They take my mishmash and1080put it all together and throw it out the other end, but it1081still looks like my stuff because it is my stuff.1082    Senator Durbin. Professor Lee, if I understand part of your1083argument here, you were suggesting that this is the age of1084innovation. Deep learning deserves special treatment. We've1085been through this argument in Congress before. Section 230 is a1086good illustration of that. We decided this fledgling industry1087called the internet just may not have a future, better be1088careful, so we exempted them from liability. Is that what you1089are suggesting?1090    Professor Lee. Not at all, Senator. My position is that we1091should pay heed to the existing Supreme Court precedence on1092fair use, which repeatedly states that fair use is a flexible1093doctrine decided on a case-by-case manner. And there is a way1094for authors to prove market harm based on a taking or the1095copying of protected elements of their works.1096    Judge Alsup said, if the authors show that there is market1097harm based on an output of this model, you could bring another1098case. And that's exactly, I think, the approach to strike the1099correct--as you mentioned earlier at the opening remarks--to1100strike the right balance between protecting copyrighted works1101and authors and protecting innovation. Even just a story in1102Emerson v. Davies recognized that not everything in a book is1103protected by copyright. Authors build on the past books to1104write new books, and that fuels creation.1105    And here, the line that Judge Chhabria and Alsup drew in1106terms of non-infringing output--or excuse me, just Judge1107Alsup--there is no copyright claim in the production of non-1108infringing works.1109    Senator Durbin. I am sorry to interrupt you, but I only1110have a minute left. It looks to me like you are shifting the1111burden to the author of the creative work when there is an1112assertion of fair use here. So Meta or others can virtually1113steal this creative product of Mr. Baldacci and others, and1114then he has the responsibility of proving that there has been1115an economic loss to him as a result of it?1116    Professor Lee. Not at all, Senator. The judges explained in1117their opinions that the--yes, the initial burden for fair use1118is on the defendant, but the defendants in both cases provided1119evidence that there was no output of infringing works. And the1120question then becomes, will the plaintiffs present contrary1121evidence? And neither judge found evidence of outputs that had1122substantially similar copies of the plaintiff's works. So the1123entire----1124    Senator Durbin. So, ultimately, the thievery, if you want1125to use that word, of the creative work is for the economic1126benefit of those who are creating the AI, is it not?1127    Professor Lee. Not necessarily. I think if the plaintiffs1128are able to prove cognizable market harm from the copying of1129their copyrighted expression, then the fair use argument is1130likely to fail for their training.1131    Senator Durbin. I am coming at it from a different angle. I1132am talking to you about why do we have AI? Why are we1133interested in AI? Clearly, it is a commercial purpose, is it1134not?1135    Professor Lee. Oh, entirely. For the AI companies, yes.1136    Senator Durbin. For the companies. So that they are1137ultimately the winners in this approach that you are taking. We1138assume we are in the world of new innovation here, and there is1139a use of someone else's creative work. The burden is on them to1140prove that they have lost money because of that piracy. But the1141ultimate winner in this is going to be the AI because if they1142escape this responsibility, they can use Mr. Baldacci's product1143and make money off of it.1144    Professor Lee. Yes, if the training is considered a fair1145use, the direct benefit would be to the AI companies. I grant1146that. But in terms of the larger national interest, it redounds1147to the benefit of the United States. If we have a priority in1148AI development, and if we are in a competition or arms race1149with China, winning the AI race by United States companies1150benefits the United States, in my view.1151    Senator Durbin. And Mr. Baldacci should be prepared to pay1152the price for that, right?1153    Professor Lee. Well, I would suggest that if it is so easy1154to generate copies of Mr. Baldacci's novels or any other1155authors, that should go in the complaint in these lawsuits. And1156some of the lawsuits do allege infringing outputs. So those are1157yet to be resolved. But my ultimate position is that we should1158not throw out the window the established Supreme Court1159precedence on how to apply fair use. It is case-by-case,1160flexible, and it balances the interests of both sides in terms1161of copyright, as well as innovation.1162    Senator Durbin. Thank you.1163    Chair Hawley. I just want to followup on this line of1164questioning, Professor Lee. When you say that it would be to1165the benefit of the United States, isn't Mr. Baldacci a citizen1166of the United States?1167    Professor Lee. Entirely. I'm not saying that Mr. Baldacci1168does not benefit from the copyright. There is another----1169    Chair Hawley. But let's take a different author, Professor1170Viswanathan. She is a citizen of the United States?1171    Professor Lee. Yes.1172    Chair Hawley. So I am just struggling to understand, when1173you say that the mass theft of their works will benefit the1174United States ultimately, you are saying that the mass theft1175and potential impoverishment of American citizens ultimately1176redounds to the good of America?1177    Professor Lee. Not at all, Senator.1178    Chair Hawley. I think you are being a little too imprecise,1179right? What you mean to say is it may benefit American1180corporations. It may impoverish American citizens, but it will1181benefit American corporations.1182    Professor Lee. Well, Senator, there is a balance to be1183struck and the courts----1184    Chair Hawley. Well, indeed, but you are waving the magic1185wand that this will benefit the United States, said we are in1186an arms race with China. I am just trying to drill down on your1187assertion. I think what you are really saying is is that the1188enrichment of certain multinational corporations that are1189incidentally based in the United States taking the works and1190personal property of American citizens is a good thing. That is1191a little bit less clear to me.1192    Professor Lee. Well, the way that I view the national1193interest, as stated by President Trump's executive order, is1194that there is a national priority in maintaining the United1195States' dominance and leadership globally in AI. And I would1196defer to the view of the AI czar, David Sacks, who said if1197there is no pathway to fair use in AI training, we will lose1198the race with China.1199    Chair Hawley. Well, you think that we should allow an1200unelected AI czar to decide what the rights of American1201citizens are?1202    Professor Lee. No, not at all. This is going through the1203courts. I would let the courts decide all of these disputes.1204And there are presently 44 lawsuits around the country, so this1205is not a time for Congress to intervene in terms of deciding1206these very difficult questions.1207    Chair Hawley. It just sounds strange to me to say that the1208United States, as a nation, is going to benefit from the mass1209violations of its citizens' rights. I thought what made us a1210nation was our common citizenship, the things that we agree on1211together, the rights that we hold in common. And your argument1212seems to be it is fine to violate those rights en masse if it1213redounds to the benefit of the Nation. I think what you are1214really saying is to the benefit of certain people in the Nation1215and their immediate interests.1216    Let me ask you about something else you said, fair use.1217    Professor Lee. Can I respond?1218    Chair Hawley. Well, just a second. I have limited time1219here. Fair use, you said, is a flexible doctrine. It is an1220equitable doctrine. And these companies aren't exactly coming1221to this with clean hands, are they? They are coming to claiming1222fair use after they have stolen Mr. Baldacci's work. They1223didn't take it from the library. They didn't license it. They1224didn't buy it. They went to a pirated illegal site and took it.1225And now they are coming and claiming the cover of equity. That1226seems kind of strange, doesn't it? Is that how equitable law1227works?1228    Professor Lee. That is the very question, the initial1229acquisition, whether that was justified as fair use. And the1230two judges disagreed on how to treat that initial acquisition1231from the shadow libraries. So I think it would be incorrect for1232us to assume that it is necessarily a violation. And the1233Supreme Court in Google v. Oracle had an opportunity to discuss1234or require considerations of bad faith in the fair use1235analysis, and it rejected that opportunity and even cited Judge1236Leval's very influential fair use article saying that fair use1237is not limited to the well-behaved.1238    Chair Hawley. Okay.1239    Professor Lee. Now----1240    Chair Hawley. We appreciate you being here, and thank you.1241You are making these arguments very gamely. That is helpful, I1242think, to have this debate. But I just want to point out that1243there is a lot of hand-waving going on here. Every time we get1244down to the nub of the question, can these giant corporations1245take the copyrighted work of individual citizens, we get1246distracted with, well, it is for the good of the country, maybe1247it is not so bad, we have an arms race on, there is an AI czar.1248Actually, I don't think it is that complicated. I think it is1249pretty simple. I think in America, we have rights. Those rights1250are what protect us. These rights are being violated. And if we1251are going to succeed as a nation and uphold our principles as a1252nation, we better darn well enforce the individual rights on1253which the nation is founded. I mean, it is just a thought.1254    Senator Welch, am I catching you off guard?1255    Senator Welch. I was kind of enjoying it.1256    [Laughter.]1257    Chair Hawley. Well, you are welcome to ask questions if you1258would like.1259    Senator Welch. I would hate to step on anyone, but1260especially a colleague Senator and the Chair of the Committee,1261you know, mid-expression of righteous outrage and indignation1262with which I am aligned, so thank you very much. Thank you. And1263I appreciate you calling this hearing because this is1264incredibly important.1265    You know, Senator Blackburn and I have a bill which is1266called the TRAIN Act, and it is trying to address this question1267of artistic content being used. And, you know, we have got a1268celebrated author here, and it would protect you. But what I1269appreciate about you being here, Mr. Baldacci, is there is a1270lot of folks who are aspiring to be David Baldacci. There are a1271lot of artists aspiring to be a Taylor Swift. And it is the1272folks who have made it that are in a position to advocate. And1273it is not, I don't think, going to benefit you, but it is going1274to benefit artists who have so much to contribute even though1275they are not yet discovered.1276    And, you know, this is the reality, and this is where I1277think the Chairman is really right. The AI companies need1278content, so they don't care where it comes from. It is just a1279voracious, insatiable appetite. And they are going to go into1280copyrighted material. We just know that. And to suggest they1281won't I think is naive. And the question and the burden here is1282that is going into copyrighted material. And the artist has the1283right to have that copyright respected.1284    The burden is that how do you know they used it? That is1285the whole point of the TRAIN Act where if there is copyright1286infringement, a reasonable assertion of that and suspicion of1287it is going to require disclosure on the part of the AI1288platform.1289    So I wanted to ask a little bit about that. And I will1290start with you, Mr. Baldacci. Do you have any suspicion that1291some of your works have been used to train AI systems?1292    Mr. Baldacci. I have been told and I have been shown a data1293base, and it's part of the--part of a class-action lawsuit1294against the AI community. And I think they've conceded that1295they've taken at least 44 of my novels and fed them into their1296large language models.1297    Senator Welch. I mean, that is astonishing. Literally, you1298have got 44----1299    Mr. Baldacci. Well, at least they didn't take them all, so1300that was nice.1301    [Laughter.]1302    Senator Welch. Just wait.1303    [Laughter.]1304    Mr. Baldacci. I know.1305    Senator Welch. And so you don't know for sure, and the only1306way you are going to find out is hopefully through this class-1307action litigation that you are part of.1308    Mr. Baldacci. Well, I certainly learned that when my son1309put in ChatGPT that ChatGPT was intimately familiar with my1310entire body of work because it was able to throw out, you know,1311plotlines that took from many of my novels, so someone had to1312feed my novels into ChatGPT. Otherwise, it could not have1313created that response.1314    Senator Welch. And we just can't allow that. You know, that1315is just really wrong. Thank you. So we are in agreement here1316that we need some reforms here to protect the artist.1317    Mr. Smith, you know, music, it is the same situation. And,1318you know, our music industry is so important. Using the word1319industry is wrong. Music is so important. It really helps1320people get a sense of who they are, it helps people connect,1321and it is across political divisions. That is what is one of1322the inspiring things about the incredible contributions that1323musicians provide to our society. And can you just explain what1324the dangers are of allowing AI models to freely train off1325copyrighted works?1326    Professor Smith. Sure. There are multiple dangers. What we1327have seen in the early piracy research is that Article I,1328Section 8, Clause 8 is actually a really good idea. Giving1329artists incentives to create actually yields more creation. And1330when artists' incomes are lowered through piracy, they have1331lower incentives to create. I think we see the same thing here,1332both directly by participating in these pirated networks, the1333generative AI companies are making it easier for other people1334to steal. But then indirectly, they're also making it harder1335for licenses to be signed. Mr. Baldacci talks about signing1336licenses, but when you sign a license with a generative AI1337company, you're signing with a gun held to your head because1338they can say, either sign what I'm offering or I'm going to go1339steal it instead.1340    Senator Welch. Well, that is the adhesion contract that1341good lawyers like Senator Hawley still remember from law school1342days. No, but explain that a little bit more because, you know,1343this is where I think all of us have some real appreciation for1344young artists. They have a vision that there is something1345inside them that they can express and that it will make a1346difference to people who hear it or people who read it. And1347they start out against their parents' will most of the time,1348right, because it is not an income-producing, promising career,1349and a lot of them don't succeed, commercial success. But they1350actually are contributing in a local community to a sense that1351helps develop our culture and helps create respect for the1352creative process and helps create respect that there are other1353things than the career path that some of us up here have1354followed where you can make a real contribution and a1355meaningful contribution.1356    So this is the concern I have about how this AI and the1357grabbing is going to make it tougher for those folks against1358great odds to keep at it. So maybe you could just, from your1359experience, talk a little bit about how it would adversely1360impact any chance they have of being able to pay their bills at1361the end of the month while they are trying to create1362inspirational music for the benefit of all of us.1363    Professor Smith. Yes, I deeply share that concern, Senator,1364and it's based on peer-reviewed academic research showing that1365creative output goes down when piracy is allowed to flourish. I1366worry that the future David Baldaccis of the world won't get1367through that hump, and we won't get to appreciate their1368creative output if we allow piracy to continue to be used to1369train these generative AI models.1370    Senator Welch. Well, thank you. My time is just about up,1371but I just want to express my gratitude to each of the1372witnesses. I didn't have a chance to speak with you, but I1373think this is an extraordinarily important issue.1374    I yield back.1375    Chair Hawley. Thank you, Senator Welch. Senator Durbin.1376    Senator Durbin. Mr. Pritt, you represent plaintiffs in a1377lawsuit against Meta that alleges copyright infringement of the1378plaintiffs' authors' works. Do you have any idea how much Meta1379as a company is valued?1380    Mr. Pritt. That's a good question. Many trillions, I1381believe.1382    Senator Durbin. Did Meta compensate any of the copyright1383owners in your case for the use of their works?1384    Mr. Pritt. No, but Meta did spend money on contributing its1385processing power to pirate from illicit websites and also to1386pay Amazon to host pirated data.1387    Senator Durbin. Which, of course, did not inure to the1388benefit of your plaintiffs.1389    Mr. Pritt. Certainly not.1390    Senator Durbin. How does the downloading and uploading of1391pirated copyright material impact the analysis of whether a1392copyright infringement could meet the mens rea requirement or1393willfulness necessary for criminal infringement?1394    Mr. Pritt. I would let the professors answer that question.1395Certainly as to willfulness in the civil copyright context, as1396the documents Senator Hawley showed, I think the answer is1397clear, that the piracy committed by Meta was knowing and1398intentional.1399    Senator Durbin. Anyone else want to comment on that? Mr.1400Lee, Dr. Lee?1401    Professor Lee. Yes, thank you, Senator. The standard of1402willfulness for criminal copyright infringement requires1403knowledge that it is illegal to engage in that particular1404copying. Now, I don't want to relitigate what Judge Chhabria1405has already ruled on, but he was given all of this evidence1406that was submitted by Mr. Pritt and his colleagues. He saw the1407comments by engineers, but he also saw comments and analysis by1408lawyers of Meta advising them on whether this was permitted or1409not under fair use law. And Judge Chhabria made a1410determination. The crime fraud exception simply didn't apply.1411    And I'm not privy to all of the analysis that Judge1412Chhabria made, but I'm assuming it was based on the question1413not being resolved, the legal question of whether accessing or1414copying from a pirated website to serve a highly transformative1415purpose is the very question raised in the lawsuit. There was1416no prior precedent that has so held that it is piracy or1417illegal, let alone criminal infringement, to do that. And that1418is the very question that Judge Chhabria ruled on. And to1419assume that it is piracy is begging the question--with all due1420respect, it is begging the question that the courts are the1421appropriate determiners of.1422    And that can be appealed, you know, and I am sure it will1423be appealed, but here the question of whether acquiring for a1424putative fair use purpose is unlawful, Judge Chhabria ruled it1425was not. It was for the fair use purpose of developing the AI1426model. I believe that is supported by the text of Section 107.1427    Senator Durbin. So Professor Viswanathan, would you like to1428comment on that?1429    Professor Viswanathan. I would, thank you so much. The very1430fact that we're talking about this kind of behavior as to1431whether or not it's criminal, right, the very fact that we're1432here talking about willful, knowing, intentional, massive scale1433training on pirated materials. Let's just step back for a1434moment from the question of whether it comes under criminal1435copyright infringement. Does it come under fair use at all? Is1436this what fair use was developed to be? Fair use, for those of1437you who don't take my copyright class, sorry about that, fair1438use is an affirmative defense. Yes, I infringed, but I did it1439for a good reason, a societally beneficial reason.1440    All right. Maybe creating a world's repository of1441generative AI companies is that, but it doesn't seem to me that1442it squares with the other things that we think of as fair use.1443What's well-established fair use? Education, criticism,1444commentary, First Amendment purposes that we consider valuable1445and necessary and that are done in good faith. I educate in1446good faith. I don't want to have to clear all those copyrights1447to educate. Okay, great, we allow you to do that.1448    That is not what's going on here. I don't want to1449relitigate the cases, Professor Lee, but Judge Chhabria was1450clearly distressed by this. And when he raised the possibility,1451as you rightly say, in dicta, that market dilution might be1452what's happening, he's saying, look, exactly what the Senator1453was talking about, flooding--what Mr. Baldacci was talking1454about, flooding the market with subpar works that substitute1455for the original works. This is not what fair use was intended1456to achieve or to facilitate.1457    And the very fact that these companies are arguing we're in1458good faith, we're doing fair use purposes, to me, this1459shouldn't even be a defense that they're allowed to raise. But1460okay, they will raise it, and it will be litigated. But boy, it1461just does not seem consonant with what fair use was ever meant1462to do.1463    Senator Durbin. Thank you. Thank you, Mr. Chairman.1464    Chair Hawley. Mr. Pritt, if I could just ask you another1465question or two about some of the evidence. We talked about1466Meta engineers saying that they realized what they were doing1467was crossing an ethical line, that they felt they shouldn't be1468doing it, but they had to cut some corners. Let me just ask1469you, did Meta ever try to hide what it was doing? Did it try to1470hide the fact that it was pirating these works?1471    Mr. Pritt. What the documents show is that in 2024, when1472Meta began to use Anna's Archive, it decided intentionally to1473not use its own servers and instead to go through Amazon Web1474Services in order to ensure that the seeding, the sharing of1475pirated works would not be traced back to Meta's own IP.1476    Chair Hawley. It doesn't sound to me like a company and1477executives that think what they are doing is above board. It1478sounds like a company that thinks that what they are doing is1479probably illegal in some manner, but they want to go on doing1480it anyway.1481    Let me just show you a couple of documents, help us1482understand what we are seeing here. These are more Meta1483engineers now, again, working on AI. We have got the first one,1484Nikolay, who says, ``not sure we can use Meta's IPs to load1485through torrents pirate content, haha.''1486    [Poster is displayed.]1487    [Laughter.]1488    Chair Hawley. I emphasize, these are their documents. I1489mean, for all of Professor Lee's--and again, I appreciate1490Professor Lee making these arguments, but for all of Professor1491Lee's comments that we are not sure if it is really pirated or1492not, they thought so. This is Meta. Meta thought so. The next1493employee, ``I'm curious to start looking at some samples, but I1494feel like we should get some clarity on what's allowed and1495how,'' smiling emoji. Nikolay again, ``haha, yes, I think1496torrenting from a corporate laptop doesn't feel right.''1497    [Poster is displayed.]1498    I mean, what are we looking at here, Mr. Pritt? I mean, is1499this an attempt to be above board and forthcoming, and, you1500know, they think everything's fine?1501    Mr. Pritt. I think that is a very difficult conclusion to1502draw from these documents. And with all due respect to1503Professor Lee, as I am still litigating the case against Meta1504on behalf of a group of authors, Judge Chhabria in that case1505specifically declined to decide whether Meta's piracy, what it1506has engaged in, in terms of the downloading, the making1507available, the making additional copies, and then sending those1508copies, over 40 terabytes of data, to other individuals, is in1509fact fair use. And no court, including the Supreme Court, has1510ever held that rank piracy is somehow fair use. And instead,1511the Supreme Court case law, still the law of the land, says1512that fair use presupposes good faith and fair dealing. I will1513leave it to you whether or not you think any of these documents1514shows good faith and fair dealing.1515    Chair Hawley. Well, let's just look at one other document1516and ask ourselves if this looks like good faith and fair1517dealing. More Meta employees, more AI engineers. ``Frank, can1518you clarify why we can't use Facebook infra''--internal--``for1519this again?'' Frank Zhang replies, ``avoiding risk of tracing1520back the seeder from a Facebook server.'' And he clarifies,1521``avoiding risk of tracing back the seeder/downloader are from1522Facebook servers.'' So here we have Meta employees saying they1523know they are pirating, they think it is ethically wrong, they1524think it is illegal, and they are actively avoiding trying to1525create a paper trail. They are trying to hide it. I mean, that1526doesn't sound like fair use to me. Does it sound like fair use1527to you, Professor Lee? I mean, do you think this is fair use?1528    [Poster is displayed.]1529    Professor Lee. I would just say I agree with Judge1530Chhabria's approach. The distribution claim is still alive in1531the case, and this aspect of the torrenting may well be1532infringement and not fair use.1533    Chair Hawley. I will just say this. If this isn't1534infringement, Congress needs to do something. I mean, if the1535answer is that the biggest corporation in the world worth1536trillions of dollars can come take an individual author's work1537like Mr. Baldacci, lie about it, hide it, profit off of it, and1538there is nothing our law does about that, we need to change the1539law. And if nothing else comes out of this hearing today, I1540hope that is it. And I hope that this is motivation to this1541body that we need to be paying attention to what is going on1542here.1543    Mr. Baldacci, you said you would rather live on a different1544planet if there was AI that could write your books. I am sure1545that that will never happen. They will never write your books.1546I want to live on a different planet if this can go on and it1547is perfectly legal. We have got to do something about this.1548    [Applause.]1549    Chair Hawley. Let me just ask you, Mr. Pritt, finally, what1550about Mark Zuckerberg in all of this? I mean, do we think that1551Zuckerberg knew about this, approved this? I mean, what does1552the evidence suggest?1553    Mr. Pritt. Certainly, the documents that have become public1554in the case explain that the decision whether or not to use1555Library Genesis, which is a notorious illicit marketplace, for1556example, for actual training as opposed to exploration was1557escalated to Mark Zuckerberg.1558    Chair Hawley. I think the judge said something to this1559effect--let's just look here if we have got it--that in fact,1560Zuckerberg was asked about it. There it is. In the spring of15612023, after failing to acquire licenses and following1562escalation up to Zuckerberg, Meta decided to just use the works1563acquired from a torrenting platform as training data. So they1564just did it anyway. They just, yes, you know, do it anyway.1565Forget it. Don't pay Mr. Baldacci. Don't pay anybody. It costs1566too much. A lot cheaper to take it for free and then make1567billions of dollars off of it.1568    [Poster is displayed.]1569    Listen, I will just conclude with this. I want to thank all1570the witnesses for their testimony. And Senator Welch, if you1571have more questions, or Senator Durbin, I am happy to let you1572ask those.1573    For my part, I just want to say, I think that this is a1574moral issue as much as anything else. I think this is an issue1575about who are we going to be as a country? Are we going to be a1576country, as it is written into our Constitution, where we1577protect the rights of our citizens? It is part of what makes us1578Americans. And we welcome the creative genius of people like1579Mr. Baldacci and the marvelous diversity of imagination and1580viewpoints and perspectives that has come to characterize our1581country. Are we going to protect that? Are we going to allow a1582few mega corporations to vacuum it all up, digest it, and make1583billions of dollars in profits, maybe trillions, and pay nobody1584for it? That is not America. That is not our country. It never1585has been.1586    Listen, I am all for the free market. I am glad Mark1587Zuckerberg can make his billions. That is fine. But not by1588running over people like Mr. Baldacci or anybody else or any1589young author who is trying to get a start or any other person,1590creative, noncreative, or just a working guy who puts something1591on Facebook. Why should all his stuff get taken? I just think1592that is wrong. I think it is morally wrong. I think, frankly,1593it is not consonant with our principles as Americans, and I1594think we can and should do better than that.1595    Senator Welch, Senator Durbin?1596    [No response.]1597    Chair Hawley. I want to thank again the witnesses for being1598here. Thanks to each of you. I know you had to travel far for1599this. And thank you again for accommodating our schedule.1600Thanks to everyone who has been here today.1601    And with that, we will stand adjourned.1602    [Whereupon, at 1:25 p.m., the hearing was adjourned.]1603    [Additional material submitted for the record follows.]1604    [GRAPHIC] [TIFF OMITTED] T1891.00116051606    [GRAPHIC] [TIFF OMITTED] T1891.00216071608    [GRAPHIC] [TIFF OMITTED] T1891.00316091610    [GRAPHIC] [TIFF OMITTED] T1891.00416111612    [GRAPHIC] [TIFF OMITTED] T1891.00516131614    [GRAPHIC] [TIFF OMITTED] T1891.00616151616    [GRAPHIC] [TIFF OMITTED] T1891.00716171618    [GRAPHIC] [TIFF OMITTED] T1891.00816191620    [GRAPHIC] [TIFF OMITTED] T1891.00916211622    [GRAPHIC] [TIFF OMITTED] T1891.01016231624    [GRAPHIC] [TIFF OMITTED] T1891.01116251626    [GRAPHIC] [TIFF OMITTED] T1891.01216271628    [GRAPHIC] [TIFF OMITTED] T1891.01316291630    [GRAPHIC] [TIFF OMITTED] T1891.01416311632    [GRAPHIC] [TIFF OMITTED] T1891.01516331634    [GRAPHIC] [TIFF OMITTED] T1891.01616351636    [GRAPHIC] [TIFF OMITTED] T1891.01716371638    [GRAPHIC] [TIFF OMITTED] T1891.01816391640    [GRAPHIC] [TIFF OMITTED] T1891.01916411642    [GRAPHIC] [TIFF OMITTED] T1891.02016431644    [GRAPHIC] [TIFF OMITTED] T1891.02116451646    [GRAPHIC] [TIFF OMITTED] T1891.02216471648    [GRAPHIC] [TIFF OMITTED] T1891.02316491650    [GRAPHIC] [TIFF OMITTED] T1891.02416511652    [GRAPHIC] [TIFF OMITTED] T1891.02516531654    [GRAPHIC] [TIFF OMITTED] T1891.02616551656    [GRAPHIC] [TIFF OMITTED] T1891.02716571658    [GRAPHIC] [TIFF OMITTED] T1891.02816591660    [GRAPHIC] [TIFF OMITTED] T1891.02916611662    [GRAPHIC] [TIFF OMITTED] T1891.03016631664    [GRAPHIC] [TIFF OMITTED] T1891.03116651666    [GRAPHIC] [TIFF OMITTED] T1891.03216671668    [GRAPHIC] [TIFF OMITTED] T1891.03316691670    [GRAPHIC] [TIFF OMITTED] T1891.03416711672    [GRAPHIC] [TIFF OMITTED] T1891.03516731674    [GRAPHIC] [TIFF OMITTED] T1891.03616751676    [GRAPHIC] [TIFF OMITTED] T1891.03716771678    [GRAPHIC] [TIFF OMITTED] T1891.03816791680    [GRAPHIC] [TIFF OMITTED] T1891.03916811682    [GRAPHIC] [TIFF OMITTED] T1891.04016831684    [GRAPHIC] [TIFF OMITTED] T1891.04116851686    [GRAPHIC] [TIFF OMITTED] T1891.04216871688    [GRAPHIC] [TIFF OMITTED] T1891.04316891690    [GRAPHIC] [TIFF OMITTED] T1891.04416911692    [GRAPHIC] [TIFF OMITTED] T1891.04516931694    [GRAPHIC] [TIFF OMITTED] T1891.04616951696    [GRAPHIC] [TIFF OMITTED] T1891.04716971698    [GRAPHIC] [TIFF OMITTED] T1891.04816991700    [GRAPHIC] [TIFF OMITTED] T1891.04917011702    [GRAPHIC] [TIFF OMITTED] T1891.05017031704    [GRAPHIC] [TIFF OMITTED] T1891.05117051706    [GRAPHIC] [TIFF OMITTED] T1891.05217071708    [GRAPHIC] [TIFF OMITTED] T1891.05317091710    [GRAPHIC] [TIFF OMITTED] T1891.05417111712    [GRAPHIC] [TIFF OMITTED] T1891.05517131714    [GRAPHIC] [TIFF OMITTED] T1891.05617151716    [GRAPHIC] [TIFF OMITTED] T1891.05717171718    [GRAPHIC] [TIFF OMITTED] T1891.05817191720    [GRAPHIC] [TIFF OMITTED] T1891.05917211722    [GRAPHIC] [TIFF OMITTED] T1891.06017231724    [GRAPHIC] [TIFF OMITTED] T1891.06117251726    [GRAPHIC] [TIFF OMITTED] T1891.06217271728    [GRAPHIC] [TIFF OMITTED] T1891.06317291730    [GRAPHIC] [TIFF OMITTED] T1891.06417311732    [GRAPHIC] [TIFF OMITTED] T1891.06517331734    [GRAPHIC] [TIFF OMITTED] T1891.06617351736    [GRAPHIC] [TIFF OMITTED] T1891.06717371738    [GRAPHIC] [TIFF OMITTED] T1891.06817391740    [GRAPHIC] [TIFF OMITTED] T1891.06917411742    [GRAPHIC] [TIFF OMITTED] T1891.07017431744                            A P P E N D I X17451746The following submissions are available at:17471748  https://www.govinfo.gov/content/pkg/CHRG-119shrg61891/pdf/CHRG-1749    119shrg1750    61891-add1.pdf17511752Submitted by Chair Hawley:17531754 Article III Project, letter......................................     217551756 Artificial Intelligence Threatens Ownership of Online Content....     517571758 Association of American Publishers (AAP), statement..............    1017591760 CreativeFuture, letter...........................................    1617611762 Motion Picture Association (MPA), letter.........................    1917631764 News Media Alliance, statement...................................    2317651766 Rumble, statement................................................    2817671768 Society of Composers & Lyricists (SCL), letter...................    3017691770Submitted by Ranking Member Durbin:17711772 Center for AI and Digital Policy (CAIDP), statement..............    3217731774 Copyright Alliance, statement....................................    3817751776 CreativeFuture, letter...........................................    1617771778 News Media Alliance, statement...................................    2317791780 Society of Composers & Lyricists (SCL), letter...................    3017811782Submitted by Senator Klobuchar:17831784 Artificial Intelligence Threatens Ownership of Online Content....     517851786Submitted by Senator Coons:17871788 Motion Picture Association (MPA), letter.........................    19