Tag: Featured

The Courts Should Embrace the Novelty of Generative AI in Copyright Law

courts

Courts can’t stick their heads in the sand to an obvious way that a new technology might severely harm the incentive to create, just because the issue has not come up before. Indeed, it seems likely that market dilution will often cause plaintiffs to decisively win the fourth factor—and thus win the fair use question overall—in cases like this. – Judge Vincent Chhabria, Kadrey et al. v. Meta –

In several posts, I have argued that generative AI (GAI) invokes novel copyright considerations on the basis that the technology has the potential to harm authorship itself, even where it may not harm specific works of authorship under traditional fair use analysis. GAI is distinguishable from any technology with which copyright law has had to contend, and if the courts will continue to guide the law to preserve copyright’s foundational principle—the incentive to create—they should recognize and even embrace the invitation to plow some new legal ground.

In the Copyright Office’s third report on artificial intelligence, one section introduces the notion of market dilution, which cites several comments including my own. Naturally, the AI industry rejects the premise that market dilution of all works, or even a certain type of work, is a valid consideration under copyright law. This argument, albeit self-interested, has some merit under traditional fair use analysis. Fair use factor four, which considers whether a specific use potentially threatens the market value of the work(s) in suit may be narrowly construed to reject the kind of generalized market harm implied by GAI.

But as the quote above reveals, Judge Chhabria in Kadrey et al. v. Meta (not even one of the strongest cases against AI developers) recognizes the novelty of this technology to undermine the foundational purpose of copyright law.  He also states, “…by training generative AI models with copyrighted works, companies are creating something that often will dramatically undermine the market for those works, and thus dramatically undermine the incentive for human beings to create things the old-fashioned way.”[1] This recognition of far-reaching harm to the “incentive” foundation for copyright addresses an even broader question than the term “market dilution” implies.

A Broader Fourth Factor Analysis

In the Copyright Office’s report, the section called Market Dilution offers guidance for a reading of fair use factor four that is broad enough to address the fact that GAI outputs can harm the overall market for the same kind of works used in training. Here, I would endorse a view that broadens the fourth factor consideration, which traditionally only looks to potential harm to the copyright owner’s exclusive right to exploit the works in suit.

As argued in other posts, and in my comments to the USCO, the courts should place considerable weight in deciding whether the use at issue furthers the purpose of copyright. My submitted comment the Office chose to highlight states: “[G]enerative AI—if it does not produce market substitutes—primarily represents potential harm to authors and future authorship. . . .[T]he 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.’”

I believe that dichotomy, novel to GAI, is precisely what the courts must resolve in order to prevent the technology from swallowing copyright law itself—perhaps especially where a given AI product does not output unlawful copies of works used in training.  The one consideration that rescues GAI products as promoting the purpose of copyright is where they are demonstrably “tools” for creators, but this goes to my overarching argument that the courts likely cannot obtain sufficient facts to discover whether the “tool” is constructive, destructive, or agnostic with regard to copyright’s purpose.

An AI tool used for ideation, for example, may further the purpose of copyright by helping the creator discover a new path from idea to protectable expression, but it is impossible for the court to assume this is the general purpose of the “tool.” The same product might just as easily be used in ways that are destructive to authorship.

For example, the vast majority of material produced and distributed will not be copyrightable due to the human authorship requirement for copyright rights to attach. Additionally, we are already seeing a large volume of AI “slop” distributed on platforms like Amazon and Spotify, and it is well established that driving traffic to garbage content is a profitable model for those willing to engage in the practice. Although a specific bucket of AI “slop,” when considered in a traditional fourth factor analysis, may not directly compete with any specific works of authorship, the courts should continue to give weight to the undeniable fact that a market flooded with “slop” does not in any sense promote copyright’s purpose and is most likely destructive to that purpose.

This view does not ignore or dismiss the creative and cultural potential of GAI as a means of expression. Many popular videos online are made with (presumably) human-authored scripts combined with AI generated AV material. That the expressions in these works will generally be unprotectable is a valid basis on which to find that the purpose of the AI product does not promote copyright’s purpose. But further, the fact that many of the creators of these works are not incentivized by copyright rights—they are motivated by the opportunity to share ad revenue with the platforms—means that these works, regardless of their qualitative value, live outside the copyright system. As such, works incentivized and enabled by a model other than copyright cannot reasonably be held to further the purpose of copyright.

In my view, these considerations look beyond the typical factor four analysis, and even beyond the ordinary concept of market dilution, to ask a fundamental question:  Can a technology built by mass copyright infringement properly make fair use of works when the product’s ultimate purpose is either destructive or irrelevant to the purpose of copyright law? I don’t think so.

Is Denial of Licensing for AI Training a Market Harm?

A recent post by Copyright Alliance CEO Keith Kupferschmid states that both Judge Alsup in Bartz and Judge Chhabria in Kadrey erred by too hastily concluding that authors are not entitled to license fees for the use of their works in AI training. On that assumption, both judges held that under factor four, the claimants could not show market harm due to the defendants’ failure to license. Kupferschmid writes:

Both judges are incorrect because they ignore the important realities that a robust emerging market for licensing of AI training material already exists. Licensing markets under the fourth factor may only be circular and non-cognizable when the market being considered is a potential licensing market and the judge is trying to determine whether that potential market is too speculative. But when there is an actual market that already exists, the circularity argument has no place and both judges were incorrect to summarily claim the argument is circular. 

Notably, Judge Chhabria, in rejecting the existence of a licensing market for AI training, cites Tresona Multimedia v. Burbank High School, but in addition to Kupferschmid’s point that a licensing market already exists for AI training, I am not sure the court’s reference to Tresona even applies. Judge Chhabria quotes from the opinion thus: “In every fair use case, the ‘plaintiff suffers a loss of a potential market if that potential [market] is defined as the theoretical market for licensing’ the use at issue in the case.” However, the next part of the opinion reads as follows:

…a copyright holder cannot prevent others from entering fair use markets merely ‘by developing or licensing a market for parody, news reporting, educational, or other transformative uses of its own creative work.’ (citation omitted)

This appears to tie the question of whether a licensing market is merely “theoretical” to a finding of whether the purpose of the use is indeed transformative. And although both the Kadrey and Bartz courts found those uses to be transformative, I believe those holdings are so tautological (i.e., lacking proper analysis) as to be ripe for significant challenge. Notably, at issue in Tresona was an educational use of small amounts of musical works—a paradigmatic fair use consideration, and one that may be as far from the implications of generative AI as we might imagine. “Further, the Warhol decision calls into question whether fair use cases like Tresona are still good law,” Kupferschmidt said to me by email.

The interplay between factors one and four, while inherent to the fair use analysis, reveals a vexing circularity in the context of GAI where the court is persuaded to find that the remarkable nature of the technology is transformative solely because the use appears to serve a “different purpose” than the works used. In addition to not fully aligning with Warhol, Judge Chhabria’s well-founded instincts about authored works “competing” with voluminous GAI works under factor four cannot be comfortably harmonized with the finding that the AI product serves a different purpose under factor one.  Clearly, if the purpose of the input material is to entertain and inform and the purpose of the “competing” output material is to entertain and inform, these are not different purposes.

The important difference, then, is that the input works are human authored, about which copyright law speaks volumes, while the output works are machine made, about which copyright law says almost nothing. In general, GAI no more adds to the productivity of copyright than the sea steadily eroding stone into an aesthetically pleasing “natural sculpture.” The courts need not attempt to foresee whether GAI will be socially beneficial or harmful but only find that in context to copyright law there are far more reasons to disfavor fair use than to favor it.

[1] I would have preferred that Judge Chhabria had not used “old fashioned,” which may be improperly read to mean “outdated” in contrast to AI generated works.

D.C. Event Shines Light on Advertisers Supporting Social Media Harm to Children

social media

When I was a kid in the 1970s and my father was a principal in an ad agency, they had the Ameritone paint account, and I remember him explaining that they were not allowed to show paint and food together in a commercial lest a child viewer be confused into thinking that paint might be edible. By contrast, a social media platform today is free to conflate child-focused material with illegal drug offers and numerous other conduits leading to serious harm or death. And it’s all swept under the rug of innovation and commerce.

Algorithms kill kids. Let’s just call it like it is at this point and stop pussyfooting around the rhetoric that social media platforms are neutral platforms for “information.” Never mind that information itself is almost a lost cause on social media, but the effects of algorithmic manipulation—even simple recommendations—can have disastrous effects for children and teens, including depression, anxiety, suicide, and accidental death. And that was before AI.

As reported last September, the accidental suicide of Nylah Anderson, age 10, was the result of TikTok’s algorithm prompting her to try the “blackout challenge,” which entails making a “game” of self-asphyxiation. In the case against TikTok for its role in leading Anderson toward the “blackout challenge,” the Third Circuit Court of Appeals articulated one of the few rational reads of the Section 230 liability shield. The court stated:

TikTok reads § 230…to permit casual indifference to the death of a ten-year-old girl. It is a position that has become popular among a host of purveyors of pornography, self-mutilation, and exploitation, one that smuggles constitutional conceptions of a “free trade in ideas” into a digital “cauldron of illicit loves” that leap and boil with no oversight, no accountability, no remedy.

Brought to You by Your Favorite Brands

Add to that cauldron the major brands whose advertising dollars unconditionally support social platforms, and that was the focus of this morning’s event held at the National Press Club. “We saw a great turnout,” says cyber-analyst Eric Feinberg, who has been engaged on ad-supported toxic social media content since 2013. More than 40 attendees filled the 40-seat room for the kick-off event designed to focus the attention of major brands on the fact that their ad dollars finance platform operations that cause serious harm and death to children and teens.

The event was organized and hosted by parents who have been working to turn personal tragedy into social change through both public policy and private action. For instance, one mother who spoke was Debra Schmill, who started the Becca Schmill Foundation after losing her daughter Rebecca to fentanyl poisoning from pills obtained with the “help” of social media. Becca’s death was the culmination in a cascade of terrible events intersecting social platforms—beginning with a rape at the age of 15 that was followed by cyber-bullying and the consequent battle with depression that led to the fatal pills obtained online. Deb Schmill is one of many parents determined to prevent other children and families from suffering similar fates.

“Women make 70% to 80% of all purchasing decisions,” Feinberg explained to me by phone after the event, “and these mothers who spoke today recognize that mothers just like them are funding social media harm to their own children.” Posting his daily mantra that “Brands are buying while kids are dying,” Feinberg has recently taken swings at McDonalds for its crossover promotion with Snapchat…

He makes a solid point. If a major brand overtly promoted the opportunity for kids to get closer to the local drug dealer, pimp, or sexual predator, parents would be outraged. But because social media is an insidious free-for-all, inhabited by good and bad actors, the worst vices are either overlooked or accepted as the cost of obtaining the virtues. But this is a false choice. Multiple defectors from these companies have made clear that the platforms bend their own rules and tweak their algorithms to promote anything that drives “engagement,” without regard to the consequences. And they assume the mainstream advertisers will keep paying without condition because they own all that engagement.

But as Meta whistleblower Sarah Wynn-Williams describes in her book Careless People, that company made an affirmative decision to target known teenage psychological vulnerabilities (e.g., body image) to promote certain products. This abuse of the technology is already unethical—a far cry from not showing paint and food on the same screen—and advertisers who knowingly exploit the “opportunity” should be held accountable by consumers. Meanwhile, as the organizers of today’s event strive to emphasize, that same algorithm exploiting the teen’s vulnerabilities will just as readily push dangerous drugs toward the child as promote a makeup product or gym membership.

By my lights, asking the advertisers to partner with their own consumers—the parents who buy their products—to pressure the platforms to adopt better practices is the very least they can do. In just a couple of months, it will be time for the ~$40 billion Back-to-School season, and as brands vie for the K-12 parents who make those purchases, they owe it to those families to pressure the digital-age media companies to stop killing kids.

Major Film Studios File Copyright Suit Against Midjourney

midjourney

“Midjourney is the quintessential copyright free-rider and a bottomless pit of plagiarism.”

Well, there it is. If you had been wondering whether and when the major studios would file a copyright infringement suit against a developer of generative AI (GAI), it finally happened on June 11. Disney and its subsidiaries, along with Universal Studios, filed a complaint against developer Midjourney alleging copyright infringement of many of the studios’ most famous and valuable intellectual properties. In the broader AI saga, the lawsuit is a big deal, though what it means for creators in general is hard to say. The studios imply that their complaint is a slam dunk, and it probably is. As the brief states:

This case is not a “close call” under well-settled copyright law. Midjourney set up a lucrative commercial service by making countless, unauthorized copies of Plaintiffs’ copyrighted works, and now sells subscriptions to consumers so that they can view and download copies and derivatives of Plaintiffs’ valuable copyrighted characters. That is textbook copyright infringement.

The 110-page complaint hardly expounds upon legal arguments and, instead, presents substantial evidence that Midjourney willfully violates the reproduction, derivative works, public display, and public distribution rights of the plaintiffs. Both as a matter of input (model training) and output of prompted materials, the studios compellingly show that their highly valuable works went into the system and that those same works come out of the system with simple prompting by users. Outputs also include expressive details like lighting and production design copied from the motion picture environments associated with famous characters.

For example, the user doesn’t just produce Darth Vader but can obtain a rendering of Vader on a Star Wars ship with lighting and production values that result in a highly detailed unlicensed derivative. The studios also allege that Midjourney itself publicly displays these outputs as a means of promoting its product’s capabilities. And finally, the studios anticipate that Midjourney will claim not to be the direct infringer and, therefore, allege that the developer is liable for secondary copyright infringement by facilitating mass infringement among its customers.

Aaron Moss, on his blog Copyright Lately, contends this case stands out among the roughly 40 active lawsuits against AI developers because the studios present so much compelling visual evidence of mass copying of protected works. While other GAI cases like Suno, Udio, and NY Times also present substantial evidence of infringing outputs, I think Moss is right that in the visual arts cases, plaintiffs rely more on the assumed infringement by means of model training more than they show obviously infringing outputs. In Disney et al., any ordinary observer can see that the characters and worlds produced are precise copies of iconic IP belonging to the studios.

Personally, I don’t see how Midjourney defends itself and, therefore, assume this case will be settled rather than fully litigated. Whatever comes of that settlement, motion picture studios generally have an interest in the development of GAI, which raises questions about independent creators and workers in the industry, as well as the future of filmmaking itself.

GAI and the Future of Filmmaking

Historically, the independent creator and the “line worker” in the industry both benefit from copyright enforcement actions of the institutional creator. For instance, when the studios go after pirate sites, the indie filmmaker benefits from a legal action she cannot afford to take, and the many crew professionals have their livelihoods protected. But with GAI, not only are the studios not seeking to shut down the developers, but they are presumably interested in the prospect of using the technology to produce motion pictures with fewer workers. This longstanding ambition of various film executives may not be attainable, but many professionals are certain that it will be attempted.

In conversations with friends in the industry, opinions vary, including those who find that GAI tools present intriguing opportunities for independent creators to produce new kinds of work at lower cost and greater speed. But at the same time, anxieties are high that GAI will mean job loss in every department of a typical production, including writers, directors, and actors.

There is no question that many motion picture jobs are threatened by GAI, but concerns that the future of filmmaking will be reduced to a few executives overseeing a roomful of programmers may be unwarranted. While GAI motion visuals are impressive and improving rapidly, it is also a shiny new toy that forecasts various cultural, financial, legal, and technological questions yet to be answered. And that’s just for film production.

My long-held view is that it was Star Wars (1977) that short-circuited the era known as the American New Wave in cinema because suddenly the prospect of the mega-franchise was much bigger than the movie itself. The end of the studio contract system resulted in greater creative independence, and a group of young directors, including George Lucas, produced what many consider a brief golden age of American motion pictures that were simultaneously box office hits, critical successes, and award winners.

When Star Wars demonstrated the potential of the film as franchise, the concurrent influx of MBA types into Hollywood amplified a new tension between film as art and film as product—i.e., the tension between filmmaker and film executive. Within that tug-o-war, it is only natural for the “suits” to want to produce as much product as they can with as little labor as possible—let alone expensive labor with opinions! Now, GAI theoretically presents that opportunity, though maybe not to the extent that many seem to either pessimistically or optimistically assume.

Usually, when the “suits” overreach with their analytics and try to predict what the market wants, the results are unimpressive. A theme I have repeated on this blog many times is that audiences want works that surprise them, not works that have been analyzed to death—and to produce fresh work takes artists, not algorithms. This rule, if it is a rule, implies a boundary that rejects the most stark predictions that major motion pictures will soon be made by three guys and a computer.

Potential Limits of GAI in Motion Pictures

Clearly, GAI output will continue to improve, and subtle renderings of naturalism will be attainable, which does imply that a whole motion picture with “human” characters can be produced without a camera or actor being involved. That this describes the future of all cinema seems unlikely, though it is notable that among the evidence presented in Disney et al., only one image is a depiction of a natural person (Mark Hamill as Skywalker), while every other infringing image is either a masked character or was originally made with pen and ink or computer illustration. Thus, the GAI’s ability to render these particular derivative works implies precisely the franchise material that could be produced without anyone building a set or rigging lights or pointing a camera at an actor. Still, there are limits.

For instance, the current Marvel franchise was primed with the first Iron Man (2008), the success of which owes a LOT to the performance by Robert Downey, Jr. This implies a caveat that films made without human artists can become what we might soon call the slop-flops of this dawning GAI era. Still, without a crystal ball or room to explore all the implications of eliminating one type of creative professional or another, a limiting factor for overuse of GAI may be copyright itself.

Because the human authorship requirement is, and will likely remain, a bedrock principle of copyright protection, GAI enables the production of a very large volume of unprotectable expression. Additionally, if two creators are using the same product, the likelihood of substantially similar, but independently created,[1] works may increase as well. In this light, creators, large and small, might want to be wary of overreliance on producing GAI material that may lack copyright protection. This concern would apply with greater force to newer characters, interpretations, and/or imagined worlds, if authorship in these works could be challenged on the basis that the expressions are the result of machine interpretation of the idea rather than human artists expressing the idea.

Personally, I would love to see the GAI genie stuffed back into its bottle because I believe that on balance the technology produces more social harms than benefits—and because the Techbros have zero credibility when it comes to ethical development or application of any of their products. But knowing that genie’s bottle has been shattered, I recognize how the technology can be used as a tool for new creative expression and am hopeful that lawsuits like this one at least push the application of GAI in that direction.

[1] Works that are independently created are, by definition, non-infringing even if they are substantial similar to other works.