Generative AI, Copyrighted Works, and the Quest for Ethical Training Practices
The legal and ethical concerns surrounding generative artificial intelligence (AI) systems being trained on copyrighted works are currently under scrutiny, with the U.S. Copyright Office conducting an Artificial Intelligence Study to address such practices. The study aims to provide insights that will assist the Copyright Office and other stakeholders in better comprehending the extent to which AI systems may be infringing protected works. Simultaneously, both a White House Executive Order and the Federal Trade Commission (FTC) are also actively exploring the potential adverse effects of AI on copyright owners and their interests, along with numerous Congressional officials holding hearings, forums, and meetings.
What do copyright holders have to say?
Hundreds of creative and copyright community members—large and small, individuals and organizations, and spanning all disciplines of creativity—have shared their perspectives in their comments submitted to the Copyright Office’s study on the use of copyrighted works to train AI systems. From concerns about loss of control over their creative works to issues related to loss of revenue, these submissions reflect the opinions of the creative and copyright community. Below are just a few of the comments received by the Office from creative stakeholders.
Association of American Publishers (AAP): Believes Congress and the Copyright Office should monitor the training of generative AI systems. AAP, an organization that represents many of the largest book, journal, and education publishers across the United States, expressed concern for its members, stating in its initial AI Study comments that “The wholesale reproduction of copyrighted works for purposes of training and developing AI systems is infringement,” that “the unlicensed ingestion of copyrighted materials for training does not qualify as fair use,” and “any framework intended to promote AI development must not diminish the copyrights of the authors and publishers whose works are essential to a free, flourishing, and well-informed society.”
American Intellectual Property Law Association (AIPLA): Believes in human created content and the essential nature of copyright. AIPLA, an organization that represents both users and owners of intellectual property, stated in its comments that, “We are compelled to note, as a fundamental value statement, that we believe human-generated creative works are irreplaceably possessed of inherent value. Art, music, dance, and storytelling are cultural universals found in all human societies throughout history. These expressive endeavors are foundational elements of human culture, [which] provide the medium for the inheritance of culture, and the shared exploration of the human condition.”
American Society of Media Photographers (ASMP) and North American Nature Photography Association (NANPA): Assert that AI technology must be implemented in a manner that respects copyright law. ASMP and NANPA, which work to protect and promote the interests of professional photographers and all visual creators, assert that, “most significantly, the need for artificial intelligence technology [must] be implemented in a manner that respects the basic principles of our copyright laws.”
American Society of Composers, Authors, and Publishers (ASCAP): Concerned that the unchecked use of AI threatens to undermine the purpose of copyright laws. ASCAP, a performing rights organization (PRO) that collectively licenses the non-dramatic public performance rights of its members’ musical compositions on a non-exclusive basis to music users that publicly perform music in virtually every communications media, asserts that, “While generative AI has the potential to enhance human creative efforts, the unchecked use of this technology threatens to undermine the very purpose of the copyright laws by supplanting, rather than supporting human creative work…the AI industry must be held responsible under all applicable laws—including existing copyright laws and state and federal legal frameworks—to ensure that it does not unfairly and illegally exploit the work of human artists, writers, and other creators.”
Authors Guild: Calls the training of AI systems using copyrighted works “self-evidently unfair.” The Authors Guild is a national non-profit association of more than 14k professional, published writers of all genres. In its comments to the Copyright Office’s AI Study, the Guild asks (and then answers) the question of why AI large language model (LLM) developers rely on pirated copies of eBooks. The Authors Guild asserts that it’s “Because the only places to get a trove of eBook texts without permission from the copyright owners (or licensing Google’s Google Books collection) are from pirate websites. The developers understood that they needed large numbers of books, but they did not want to pay to license them.” The comments go on to assert that “This practice must be brought to an end…Not only is it self-evidently unfair, but if unrestrained, will put many human writers out of work and devalue our literature and culture.”
Copyright Clearance Center (CCC): Asserts that “AI development must be paired with an appreciation of and respect for creators and copyright.” CCC, an organization that has more than 40 years of experience in copyright and information management solutions, noted in its AI Study comments that, “…AI systems have incredible potential to support our society and economy in ways both familiar and yet unknown. To fulfil this potential, AI development must be paired with an appreciation of and respect for creators and copyright. Copyright is an engine of innovation, a key part of economic activity, and incentivizes the creation of foundational materials upon which AI is often built. Support for copyright is crucial to our culture, science, jobs, and the advancement of AI itself.”
Digital Media Licensing Association (DMLA): Wants to ensure that AI models respect copyright and that content is lawfully obtained. DMLA, which represents the interests of entities that primarily license still and motion images to publishers, the media, advertisers, designers, and others, asserts that “[it] supports the potential and opportunity that generative AI models offer but wants to ensure that the models respect copyright, that training content is lawfully obtained, that there is transparency in the content that is used to train the models, including that records are maintained and accessible in order to identify content that is used in training, and that the creative community continues to receive benefit if their works are used in the creation of generative.”
Directors Guild of America (DGA): Believes in the need to safeguard the creative vision of directors. DGA, which represents the interests of more than 19,000 directors and members of the directorial teams who create feature films, television programs, commercials, documentaries, news, and other motion picture productions, noted in its AI Study comments that, “It is essential for policymakers and the courts to strike a balance… by developing meaningful and enforceable guardrails to prevent technological advancements from undermining intellectual properties, job opportunities, and artistic integrity…The technological boundaries of [generative] AI are virtually limitless, emphasizing the need to safeguard the integrity and the singular creative vision of a director.”
Entertainment Software Association (ESA): Supports the use of AI and copyright laws and protections. ESA, which represents the video game industry, is no stranger to the use of AI in gaming. As noted in ESA’s comments,“Artificial intelligence technology has been deployed in games for over two decades as useful tools for a variety of purposes, such as background and terrain generation, processing or analysis of data within a game, or quality control.”
However, ESA further noted that “Copyright protection is vital to the innovative AI technologies incorporated into video games,” and it urged the Copyright Office to “make legal and policy recommendations to Congress that will incentivize creativity [and] encourage copyright registration,” along with the advancement of generative AI technology.
Getty Images: Supports “responsibly developed and properly licensed AI models.” Getty Images, a visual content creator with more than 800k global customers, and the co-developer of Generative AI by Getty Images (which is trained exclusively on Getty Images content and data), submitted comments to the Copyright Office’s AI Study in support of “responsibly developed and licensed AI models.” However, Getty’s comments also go on to note that “There are significant risks to consumers, creators, and the public interest when developers of AI Systems and AI Models exploit copyrighted content without permission from the relevant copyright holders.”
Vince Gilligan: Asserts that AI companies are currently “hawking guessing machines.” Vince Gilligan is the creator of the hit TV series “Breaking Bad.” In his comments to the Copyright Office’s AI Study, Gilligan states, “I find the phrase ‘artificial intelligence’ to be problematic. To me, it’s false advertising. It implies that the current state of the art is a technology that can independently think and create—when in fact it can’t and doesn’t. Someday, true generative artificial intelligence may come to exist. But right now, what companies such as OpenAI, Google and Meta are hawking are only elaborate guessing machines…I’ll admit, it’s an impressive trick. But it’s not creation. It’s not storytelling…Meanwhile, the large language models these companies use to work their magic are made up of hundreds of millions of pages of novels, textbooks, screenplays, magazine articles, social media posts, limericks, clipped recipes, you-name-it. I’m sure every word of ‘Breaking Bad’ has been jammed in there somewhere. Only, I don’t remember giving anyone the okay to do that.”
Graphic Artists Guild: Says it is “critical that copyright remain true to its purpose: to incentivize creators by assigning to them the exclusive rights to their works.” The Graphic Artists Guild, a trade association representing the interests of illustrators, designers, web developers, animators, and other visual artists that rely on licensing, notes in its AI Study comments that, “The major concerns [for our members] are the use of visual artists’ copyrighted images, without consent or compensation, in [AI] training datasets; competition from the vast quantities of visual content created by AI image generators; the generation of visual works, which resemble their original works in the outputs; and the ability of users to [copy] a visual artist’s unique style. To protect the livelihoods of professional graphic artists, it is critical that copyright remain true to its purpose: to incentivize creators by assigning to them the exclusive rights to their works. Policies around generative AI must consider first and foremost the interests of the human creators, without whom this technology would not exist.”
Motion Picture Association (MPA): Believes the U.S. Copyright Office has “an important role to play in ensuring a careful and considered approach to AI and copyright.” The leading global advocate of the film, television, and streaming industry, believes that the Copyright Office will play a pivotal role in what unfolds regarding copyright and AI, noting that “MPA’s overarching view, based on the current state, is that while AI technologies raise a host of novel questions, those questions implicate well-established copyright law doctrines and principles. At present, there is no reason to conclude that these existing doctrines and principles will be inadequate to provide courts and the Copyright Office with the tools they need to answer AI-related questions as and when they arise. The Copyright Office has an important role to play in ensuring a careful and considered approach to AI and copyright.”
National Music Publishers’ Association (NMPA): Urges U.S. Copyright Office to “err on the side of protecting human creators.” NMPA, the trade association representing American music publishers and their songwriting partners, urged “proactive protections for creators” by noting that, “[T]he development of the generative AI marketplace is marked by breathtaking speed, size, and complexity. Hindsight may well prove that there is no hyperbole in saying that generative AI is the greatest risk to the human creative class that has ever existed…Even more alarming is that we do not know how long the window is to act before it is too late. We therefore implore the [U.S. Copyright] Office to support proactive protections for human creators and, where there is uncertainty, to err on the side of protecting human creators.”
News/Media Alliance: Finds pervasive unauthorized use of publisher content to power generative AI technologies. The News/Media Alliance, which represents more than 2,200 diverse U.S. news and magazine publishers—recently published a white paper and a technical analysis, along with submitting comments to the Copyright Office on the use of publisher content to power generative artificial intelligence technologies. Per its comments, the News/Media Alliance asserted that, “Together, the [study comments, white paper, and technical analysis] document the pervasive, unauthorized use of publisher content by AI developers, the impact this may have on the sustainability and availability of high-quality original content, and the legal implications of such use.” Based on its white paper and research, the News/Media Alliance requested that the Copyright Office publicly clarify that use of publishers’ content for commercial generative AI training and development is likely to compete with—and harm—publisher businesses, “which is disfavored as a fair use.”
Professional Photographers of America (PPA): Believes there must be accountability surrounding photos used to train AI systems. With nearly 35k members, PPA is one of the world’s largest nonprofit organizations that serves professional photographers. In PPA’s comments to the U.S. Copyright Office, it stated, “Our knowledge, understanding, and view of our natural world will diminish as creators are no longer incentivized to create…Our wonder for the world will be replaced with the wonder for whether anything we see is real or just more AI. In this almost certain eventuality, all of humanity loses. To the extent that AI systems generate images that are substantially similar to the creative work of photographers, there must be accountability. They cannot be allowed to steal photographers’ work and then compete with those photographers using that same work.”
Recording Industry Association of America (RIAA) and the American Association of Independent Music (A2IM): They are not seeing AI implemented responsibly or ethically. RIAA and A2IM represent individual music makers and music organizations. In filing their comments together for the Copyright Office’s study, the organizations noted, “AI can be enormously beneficial when it is implemented in a responsible, respectful, and ethical manner. Like every new technology, AI will undoubtedly push creative boundaries and help shape recording artists’ visions and expand their commercial reach. We embrace AI’s potential as a tool to support human creativity, provided that it is not used to supplant human creativity. As exciting as AI is, … we are not seeing it implemented in a responsible, respectful, and ethical manner. In particular, the unauthorized ingestion of our members’ copyrighted works for purposes of training generative AI systems amounts to copyright infringement on a massive scale and causes significant economic harm to our members and their sound recording artists.”
SAG-AFTRA: Concerned unauthorized AI use will discourage future human creativity and expression. SAG-AFTRA, the nation’s largest labor union representing working media artists with more than 160,000 members that include actors, recording artists, broadcasters, and many others, noted in their comments that, “The number of AI technologies and the quality of AI-generated content has increased exponentially in a very short time. Unchecked, we are rapidly barreling toward a future where creators will have to compete with unauthorized versions of themselves or their works. This is already starting to happen in the creative fields; it is unfair, and it will discourage future human creativity and expression.”
What will happen next?
The initiatives that will follow the Copyright Office’s AI Study remain unknown, and a similar degree of ambiguity surrounds the ongoing efforts of Congress, the White House, and various government agencies that are striving to navigate the intersection of AI and copyright issues. However, one consistent theme emerges from the comments submitted to the Office’s AI Study by members of the creative community. Across the board, there is a collective assertion that more ethical and transparent training practices must be employed by AI companies. This includes obtaining permissions from—and providing compensation to—the rightful owners of the content used.
If you aren’t already a member of the Copyright Alliance, you can join today by completing our Individual Creator Members membership form! Members gain access to monthly newsletters, educational webinars, and so much more — all for free!