Category: Artificial Intelligence (AI)

Comments to the Copyright Office on Artificial Intelligence

Below are the responses I submitted to selected questions in the U.S. Copyright Office Notice of Inquiry and request for comments on artificial intelligence.

8.1. In light of the Supreme Court’s recent decisions in Google v. Oracle America and Andy Warhol Foundation v. Goldsmith, how should the “purpose and character” of the use of copyrighted works to train an AI model be evaluated? What is the relevant use to be analyzed? Do different stages of training, such as pre-training and fine-tuning, raise different considerations under the first fair use factor?

In my view, neither case is helpful to a putative AI-developer defendant regarding the first factor question being asked. Under Warhol, there is no colorable defense that the purpose of AI training is to achieve “critical bearing” on the works used, and it is difficult to imagine how most, if any, developers would make such a claim. In Oracle, the reimplementation of APIs for the development of new computer programs is highly distinguishable from, for instance, copying a billion images in their entirety to “train” a machine to generate images. Further the Court cautioned that Oracle is narrowly tailored to computer programs as the copyrightable works in question.

8.5. Under the fourth factor of the fair use analysis, how should the effect on the potential market for or value of a copyrighted work used to train an AI model be measured? Should the inquiry be whether the outputs of the AI system incorporating the model compete with a particular copyrighted work, the body of works of the same author, or the market for that general class of works?

This is one example in which generative AI can upend copyright doctrine. Even where a use may involve millions of works (e.g., Google Books), the fourth factor considers potential harm to the market for the works, whereas generative AI—if it does not produce market substitutes—primarily represents potential harm to authors and future authorship. While it is beyond the scope of copyright to protect creative jobs against technological changes per se, the consideration in the context of “training” should be expansive and doctrinal—namely that a potential threat to “authorship” cannot, by definition, “promote the progress” of “authorship.” Therefore, the fair use defense should be unavailable to the developer in this context. Where the AI does produce market substitutes, case law should be sufficient, including the fair use factor four inquiry.

10.3. Should Congress consider establishing a compulsory licensing regime? If so, what should such a regime look like? What activities should the license cover, what works would be subject to the license, and would copyright owners have the ability to opt out? How should royalty rates and terms be set, allocated, reported and distributed?

No. With the development of streaming platforms, compulsory licensing devastated the livelihoods of songwriters. AI development is still nascent, and we cannot predict how the market will change in the future. Legislation of this nature is likely to be short-sighted and may lock in regimes that fail to serve authors.

  1. In order to allow copyright owners to determine whether their works have been used, should developers of AI models be required to collect, retain, and disclose records regarding the materials used to train their models? Should creators of training datasets have a similar obligation?

Yes. Both parties should be obligated to collect and maintain records to foster transparency. That said, considering the lack of good faith shown by the tech sector in complying with regimes like the DMCA, provisions of this nature should contain actual penalties for failure to comply. Any proposal based upon a civil liability shield in exchange for compliance should be dismissed as a non-starter and a waste of time.

  1. Under copyright law, are there circumstances when a human using a generative AI system should be considered the “author” of material produced by the system? If so, what factors are relevant to that determination? For example, is selecting what material an AI model is trained on and/or providing an iterative series of text commands or prompts sufficient to claim authorship of the resulting output?

The threshold for copyrightability must still be a “modicum of originality” contributed by the human, whether the tool used is a camera, keyboard, paintbrush, or, perhaps, an AI-based application. It is certainly possible for a human using a generative AI to be the “author” of the work, though, I believe, not the developer of the AI model or system. Merely selecting the materials used to “train” an AI cannot be considered “authorship,” let alone “authorship” of what may be tens of millions of outputs. Further, because it appears that most models require the ingestion of tens of millions of works in order to “learn,” this volume of collection by means of, for example, internet scraping is too indiscriminate to be considered “selection.”

  1. Is legal protection for AI-generated material desirable as a policy matter? Is legal protection for AI-generated material necessary to encourage development of generative AI technologies and systems? Does existing copyright protection for computer code that operates a generative AI system provide sufficient incentives?

My answer to the first two questions is No, and, therefore, the third question is moot. In general, it is not desirable to attach copyright rights to AI-generated material any more than it is desirable to vest civil rights in robots. As stated above, if individual humans use certain AI-based tools to create works of expression, the use of these tools should not automatically disqualify the entire work from copyright protection.

But we must be cautious about vesting copyright rights in enterprise-scale, corporate production of works by, for instance, an AI-developer/producer. Beyond posing a threat to the careers of creative professionals (and to the cultural value of creative work), at a certain point, the application of copyright law itself may become irrelevant and/or unconstitutional. For instance, if generative AI were to foster an oligopoly of developer/producers, it is conceivable that copyright enforcement would become meaningless. Imagine the chaos (or futility) arising from a claim that AI-Developer Alpha allegedly infringed the work of AI-Developer Beta. Such a scenario raises difficult questions of standing and, as noted below, may frustrate the substantial similarity inquiry to the point of irrelevance.

Meanwhile, if these potential outcomes result in shrinking the population of working authors and the diversity of works, this would be anathema to the constitutional purpose to “promote progress.”

  1. Does the Copyright Clause in the U.S. Constitution permit copyright protection for AI-generated material? Would such protection “promote the progress of science and useful arts”? If so, how?

While patent protection for AI systems may promote the “useful arts,” copyright protection for AI-generated works does not inherently promote “science” (a.k.a.) new creative expression. Again, if generative AI is likely to reduce the number of working “authors” in the U.S., this is offensive to the Progress Clause in Article I. That the American Framers could only conceive of human “authors” is not just a technicality of history. Whether in 1787 or today, law, like art, is a human construct that serves no purpose beyond human experience. As I have stated a few times here, it is contradictory to believe that one can promote “authorship” while obviating the role of “authors.” That the Progress Clause could be interpreted to encompass this result defies textual, doctrinal, and historical reason.

  1. Can AI-generated outputs implicate the exclusive rights of preexisting copyrighted works, such as the right of reproduction or the derivative work right? If so, in what circumstances?

Yes, and it’s happening right now. Users of generative AI are producing famous—and famously protected—reproductions and derivative visual works of, for instance, Marvel characters. While copyright owner Disney may not elect to take legal action against, for instance, parties sharing these outputs on social media, there is nothing about the use of AI in these examples that militates against finding that the copies and derivatives are infringing. In fact, it seems certain that if these outputs were to be used commercially, the inevitable litigation would be short work for the court.

As to whether the AI developer may be liable for these copies and derivatives, it seems straightforward to find that if, for instance, “Daredevil” was input for training and “Daredevil” was later output by the system, then the developer may be liable for both direct and secondary copyright infringement. Direct copyright infringement in violation of §106(1) occurs during input, and secondary infringement arises due to the developer’s failure to prevent the infringing work from being output.

  1. Is the substantial similarity test adequate to address claims of infringement based on outputs from a generative AI system, or is some other standard appropriate or necessary?

Combined with 24 below.

  1. How can copyright owners prove the element of copying (such as by demonstrating access to a copyrighted work) if the developer of the AI model does not maintain or make available records of what training material it used? Are existing civil discovery rules sufficient to address this situation?

The doctrines of “substantial similarity” and “access” may both be challenged by generative AI. First, if a given system is prompted to produce common or popular probability outcomes, it may generate thousands or millions of similar works, all of which are potentially non-infringing under the doctrine of “independent creation.” For example, the AI user who is unfamiliar with the work of Karla Ortiz may inadvertently produce a work that is “substantially similar” to one of Ortiz’s images, but the copy may be said to be “independently created” by that individual. The difficulty is novel because “independent creation” is historically anomalous whereas AI could make it rampant.

This also goes to the question of “access” and liability for that “access.” The AI developer undoubtably has “access” to any work fed into its model, but it seems unlikely the user of the AI can be said to have had “access” in this context. Then, because of the increased likelihood of multiple, “independently created” similar works, proving “access” may be moot with regard to the liability of the individual user of the AI.

  1. If AI-generated material is found to infringe a copyrighted work, who should be directly or secondarily liable—the developer of a generative AI model, the developer of the system incorporating that model, end users of the system, or other parties?

Assuming the user of the AI knowingly made the infringing work, he/she is liable for direct infringement. Secondary liability for any developer may arise if the allegedly infringing work is a copy, or is “substantially similar,” to a whole work that was fed into the model or subsequent application.

25.1. Do “open-source” AI models raise unique considerations with respect to infringement based on their outputs?

I fail to see why “open source” would alter the consideration of an alleged infringement.

  1. If a generative AI system is trained on copyrighted works containing copyright management information, how does 17 U.S.C. 1202(b) apply to the treatment of that information in outputs of the system?

As in pre-AI considerations, §1202(b) denies the AI developer an “innocent infringer” defense.

  1. Should the law require AI-generated material to be labeled or otherwise publicly identified as being generated by AI? If so, in what context should the requirement apply and how should it work?

A requirement to identify AI-generated material likely addresses topics outside the scope of copyright law. Presumably, the public would be best served by labels used to mitigate fraud and other forms of misinformation, and the complications arising from that intent are best left to the FTC, Congress, and agencies other than the Copyright Office. Even where AI may be used to create forgeries, this is already criminal conduct, and copyright plays little or no role. But unless a creative work is illegally presented as the work of a named artist, there is no compelling interest per se in notifying the public that a work, or part of a work, was generated by AI.

  1. Should Congress establish a new federal right, similar to state law rights of publicity, that would apply to AI-generated material? If so, should it preempt state laws or set a ceiling or floor for state law protections? What should be the contours of such a right?

Again, this not a copyright matter, but a federal ROP may address some of the already rampant, unethical uses of AI where the potential harm to both the infringed party and the public is significant. The 25 state ROP laws do not address, for instance, the potential harm caused by AI’s capacity to generate “in the style of” works, especially in the commercial market.

If ROP law is expanded, it should 1) apply to all persons, not just celebrities; 2) anticipate and remedy AI-enabled harms stemming from misappropriation of likeness for purposes other than commercial advertising; and 3) not restrict expressive uses of AI-generated likeness for purposes (e.g., biographical films) that fall within the scope of protected speech.

In a world in which all media travels the globe instantly, a federal ROP statute would seem to be the only sensible framework in which to address the myriad potential harms. As for preemption, this may raise the cost of potential litigation by eliminating the option of a state filing, but further study into this as a matter of civil procedure is required.

  1. Are there or should there be protections against an AI system generating outputs that imitate the artistic style of a human creator (such as an AI system producing visual works “in the style of” a specific artist)? Who should be eligible for such protection? What form should it take?

This concern could be addressed in a new, federal ROP statute while leaving undisturbed the doctrine that copyright does not protect “style.” While an amendment to the Copyright Act akin to VARA could be written to encompass a narrow protection for “style,” the intent seems better suited to ROP. Additionally, it may be easier and more effective to write a new law for the purpose of federal ROP than to amend the Copyright Act, especially when the U.S. is not a moral rights jurisdiction vis-à-vis copyright. Finally, it should be noted that, if protection of this nature were enforceable, it may create new licensing opportunities for artists and prospective commercial users.

As to application of the law, again, forgery is covered by the criminal code, but the most likely harm would seem to be commercial use of “in the style of” works in a manner that may implicate the artist’s reputation and/or deny her a commission or a licensing opportunity. Still, if such a right were to be established, exceptions would be required so that what we might call “reminiscent of” is distinguishable from “in the style of.” This is a highly subjective consideration that may draw lessons from “substantial similarity” doctrine, even if the new right does not sound in copyright. For instance, prompting an AI for a work “in the style of [named artist]” may be analogized to the principle of “access.”

  1. Please identify any issues not mentioned above that the Copyright Office should consider in conducting this study.

In regard to disclaiming AI-generated material in a registration application, the current guidelines are likely to confuse applicants and overburden examiners who have neither the resources, nor necessarily the expertise, to engage in assessments normally left to the courts. Although it is understandable that the Office wishes to preserve the “human authorship” doctrine, asking an applicant how a work was made is a significant shift that should not be taken lightly.

Applicants are going to submit myriad statements which are either unfounded in law, or which beg for what amounts to a “substantial similarity” test on the part of the examiner. This may strain Office resources and potentially cost applicants additional fees. Instead, if the Office were to add a checkbox at the Certification stage of the application, asking whether the deposit copy(ies) contain any AI-generated material, the applicant will be given the opportunity to make a truthful statement subject to §506(e) while leaving the question of separating the AI material from the human authorship to the courts, as it should be.

Opportunity Costs (and with AI it may cost a bunch)

Lately, one reads a lot of statements with the preamble “Artificial intelligence presents opportunities and challenges…” But is this the right way to frame the conversation? Because if we’re talking about creative professionals and their industries, it is probably more accurate to say that generative AI presents clear threats and some opportunities. Although we are trying to predict future outcomes, and many expectations about AI (good or bad) may not come to pass, if generative AI is an existential threat to potentially millions of creative professionals while offering opportunities for a few, then it is wrong to begin the discussion as if opportunity and challenge are balanced forces.

Take, for example, the tentative agreement reached between the Writers Guild of America (WGA) and the motion picture producers, which includes the following provisions regarding the use of artificial intelligence:

  • AI can’t write or rewrite literary material, and AI-generated material will not be considered source material under the MBA, meaning that AI-generated material can’t be used to undermine a writer’s credit or separated rights.
  • A writer can choose to use AI when performing writing services, if the company consents and provided that the writer follows applicable company policies, but the company can’t require the writer to use AI software (e.g., ChatGPT) when performing writing services.
  • The Company must disclose to the writer if any materials given to the writer have been generated by AI or incorporate AI-generated material.
  • The WGA reserves the right to assert that exploitation of writers’ material to train AI is prohibited by MBA or other law.

These conditions prove the point in that they primarily seek to mitigate the threat of AI while opening a narrow and conditional window for the opportunity to use AI. Safeguards like these are necessary because it can be assumed that producers and show runners will be tempted by the prospect of paying fewer writers to “collaborate” with generative AI to produce scripts. But even if that approach were to prove effective (and there are reasons to think it would not), a writers’ room of, say, two instead of ten is not necessarily an opportunity. And perhaps not even for the show runners for very long.

Thinking solely about the U.S. economy, those laid-off writers would represent eight middle-class jobs lost—eight people who would curtail, if not cut off, their entertainment expenditures while they take the “opportunity” to ply their skills in other fields that may also be shedding jobs due to AI. If AI were to reduce the workforce in the entertainment industry alone, it would suck but could potentially fall within the principle of creative destruction. But if AI decimates work across multiple sectors at the same time, then products, including TV shows and movies, will lose customers, thereby nullifying those short-term savings gained by laying off those eight writers.

Meanwhile Creative Work Would Start to Suck

Beyond considering whether generative AI is an opportunity in cold, economic terms, it is hard to imagine outcomes that do not either diminish the cultural value of creative expression itself or trigger a rebellion against AI-generated material and dash the ambitions of the tech developers. In this regard, the “democratization of creativity” is a woefully ignorant goal as well as a dishonest talking point.

The promise that generative AI will “democratize creativity” should be read in the same light as Big Tech’s promise to “democratize information,” which has proven disastrous for democracy. Just as searching the web for “information” does not make the individual a journalist, instructing a generative AI to render ideas into expression does not make the individual an artist. And just like we continue to founder in a sea of disinformation, there is no broad, social value in “democratized” art any more than there is a market for children’s drawings tacked to a million refrigerators. If everyone is an artist then nobody is, and the value of creative expression diminishes accordingly.

That the creative process can be reduced to an algorithm which can learn how to write, draw, paint, etc. cannot be wholly denied when generative AIs are already doing these things and will presumably get better at doing them. However, the expectation that generative AI can or should displace artists may be the apotheosis of the TechBros’ enduring cynicism about the value of individual creators. In the trenches of the “copyright war,” creative professionals have been accused of being self-important, greedy, rent-seeking, whiners unwilling to get real jobs. And now that Big Tech is releasing tools that promise to obviate the need for creators, the newest hashtag claims that professional artists enjoy a #CreativityPrivilege that will finally be disrupted. In this context, generative AI can be seen as tech’s nuclear strike in the copyright war to prove once and for all that “original expression” is an illusion and, therefore, that any rights associated with original expression are a mythical construct that must be abandoned.

This impliedly jealous relationship with artists is an extension of the problem that the tech-utopian, anti-copyright crowd has never quite understood what artists do or why they do it. For instance, artistic output is not solely the result of interest plus training. Many great artists never receive formal training, and many need to escape formal training to find their own voices. Every artist will eventually, if not continually, go through a process of learning and unlearning various “rules” to make the craft their own. It may be a cliché to think of the artist as suffering or broken, but it is certain that the artist is sensitive to the world in a way that she is moved to respond through expression. And these are just some of the unpredictable human qualities that no computer can emulate with the math of probability outcomes.

Although it is plausibly argued that a creative-minded individual might have a disability which AI can help overcome, citing this hypothetical to justify the “democratization” narrative comes with a few caveats including:  1) enabling the few does not justify displacing the many; 2) if AI devastates the professional, creative ecosystem, the newly enabled artist can only be a hobbyist among millions of other hobbyists; and 3) if anyone believes the billion-dollar investments in generative AI were made with the intent to help someone with cerebral palsy become a painter, I’m calling billion-dollar bullshit. That may be a positive effect, but it is not the purpose of these machines.

Could the Models Simply Fall Down?

If generative AIs were to displace enough professional artists, it is possible that entropy will demand that the models exhaust their capacity for new outputs—let alone outputs that are of any interest or value. If we remove, say, one million working artists from the equation over the next few years, what will continue to feed the training models? Is the “sum of all human output” as of today sufficient to enable a generative AI to produce infinite, relevant expressions indefinitely? Maybe. But not necessarily.

Because artists are people who respond to the world through expression, timeliness and context matter a great deal. There are many reasons–from aesthetics to subject matter–why theater of the 19th century or television programs of the 1980s or ad campaigns of the 1960s are anachronistic to a contemporary audience. Yes, certain works endure or become freshly relevant as remakes because human experience is, in part, cyclical. But it is the artist’s sensitivity to the contemporary world that makes those connections, and the process of synthesizing that into creative expression is often instinctual as much as it is intellectual.

Yes, artists recycle and build upon prior works, but the relevance of a new expression at a given time and place requires a connection with audience that, again, is not merely the result of a probability outcome. This anticipates the likelihood that a lot of AI-generated work will be good enough but not necessarily good—a concern that directly affects the market for commercial art where many creators make a living.

For example, the stock music market for commercial use is built on a network of composers with the skills to produce a variety of tracks based on familiar and, often popular, music. If generative AI can adequately produce similar tracks by cutting out the human composer, the market for many composers is in peril. But again, if AI were to kill off or dramatically reduce new, human composition, it is conceivable that the “composition machine” might eventually fizzle out as it tries to burn the same fuel over and over.

No doubt, artificial intelligence will seed new opportunities, though I maintain that these are in fields other than the production of creative work. If the digital revolution in the creative market has taught us anything, it is that these technologies are generally an opportunity for owners of the tech at a tremendous cost to professional creators. Without the right safeguards, AI could exacerbate this trend in ways that will cost everyone.

Photo by: robcaven

Chabon v. Chatbot:  About those ‘Shadow Libraries’

As many readers already know, another class-action lawsuit was filed on September 8 against OpenAI by book authors Michael Chabon, David Henry Hwang, Matthew Klam, Rachel Louise Snyder, and Ayelet Waldman on behalf of all authors similarly situated. The allegations are almost identical to the complaints in other class-action suits against various AI companies. I won’t repeat what I have already written about each allegation, but once again, I predict that if the court does not find unlawful reproduction in transient copies necessarily made in RAM, Open AI will likely prevail. Once again, this complaint alleges that the GPT model itself is an unlicensed “derivative work” of the entire corpus of books fed into it, but this does not seem to be a well-founded implication of the derivative works right under copyright law.

But one aspect of this complaint (as well as Tremblay et al.) is that Open AI is alleged to have obtained part of its database from known pirate repositories. In reference to one of the datasets used to train Chat GPT, the Chabon complaint states, “the only ‘internet-based books corpora’ that have ever offered that much material are infamous ‘shadow library’ websites, like Library Genesis (“LibGen”), Z-Library, Sci-Hub, and Bibliotik, which host massive collections of pirated books, research papers, and other text-based materials. The materials aggregated by these websites have also been available in bulk through torrent systems.” So, is the act of exploiting illegally obtained materials in this manner a violation of law?

Certainly, the Copyright Act does not address the issue. There is language about “lawfully made” copies in the context of first sale doctrine and certain exceptions for libraries. The only two uses of the words “lawfully obtained” in Title 17 pertain to acquisition of a computer program and permissible circumvention of technical protections for research purposes. So, nothing in the Copyright Act makes Open AI’s scraping “shadow libraries” an infringing act on its own, and there is no language in §107 on fair use that refers to lawfully making or obtaining material(s). This would be anathema since a fair use defense implies an unlicensed use.

Still, it seems wrong (probably because it is) to profit by exploiting another party’s unlawful possession of valuable materials. Under the criminal code (Title 18 §2315), it is a “federal offense to receive, possess, barter, sell, or dispose of stolen property with an aggregate value of $5,000 or more if the property crosses state lines.” The statute refers to physical property and not to exploiting databases full of pirated material. But if an AI developer knowingly exploits repositories replete with unlicensed copies of works, doesn’t that sound like it should be illegal?

This discussion reminds me a little bit of the rationale for the Protecting Lawful Streaming Act of 2020, which elevated the unauthorized public performance of works via streaming from a misdemeanor to a felony. After years of debate—and allegations by anti-copyright groups that felony streaming would be disastrous—Congress recognized that unlawful streaming is effectively a digital-age version of mass bootlegging physical copies, which had long been a felony. In fact, streaming is worse because it can reach a much larger black-market than any bootlegger distributing physical products ever could.

So, under a similar rationale by which Congress recognized that streaming digital repositories of unlicensed works is a felony, perhaps lawmakers might broaden the intent of Title 18 §2315 to prohibit mass exploitation of digital warehouses full of illegal copies of copyrighted works. Certainly, these warehouses contain materials with aggregate values in the tens of millions of dollars. Hence, any party that knowingly exploits these warehouses for financial gain might reasonably be liable under the criminal code.

Authors and artists are justifiably angry that their works are being used without permission to train generative AIs. And the fact that Chat GPT was allegedly trained in part with corpora of literary material acquired and stored by media pirates is salt in the wound to say the least. I don’t know what, if any, legal remedies might be proposed, but I am confident that it is generally wrong to profit from the intentional use of ill-gotten goods.

Photo by: onephoto