BEFORE THE OFFICE OF SCIENCE AND TECHNOLOGY POLICY
Docket No. OSTP-TECH-2025-0067
The Digital Media Licensing Association (DMLA) appreciates the opportunity to submit the following comments in response to the Request for Information (RFI) published by the Office of Science and Technology Policy (OSTP) in the Federal Register on September 26, 2025. The RFI seeks input from interested parties on identifying existing Federal statutes, regulations, agency rules, guidance, forms, and administrative processes that may unnecessarily hinder the development, deployment, and adoption of artificial intelligence (AI) technologies within the United States.
The DMLA is a non-profit, non-partisan trade association founded in 1951 that represents the interests of content creators, digital media producers, distributors, and licensors. Our membership includes thousands of professionals across the visual content industry, from individual photographers and illustrators to major stock agencies, technology innovators, and AI developers. For more than seven decades, DMLA has worked to establish business standards, develop best practices, and advocate for copyright protection, privacy rights, fair licensing practices, and ethical AI development. DMLA members license millions of images, videos, illustrations, vectors, audio, and other creative content globally each day, powering everything from news media and educational materials to corporate communications, advertising campaigns, and entertainment products.
DMLA commends OSTP and all participating federal agencies for their coordinated effort to ensure that America’s leadership in AI innovation remains strong. We have been actively engaged in discussions regarding AI and copyright with our members, industry partners, elected government officials, and government agencies.
DMLA supports the responsible, lawful, and ethical development and use of AI technologies, along with a thriving and sustainable AI economy. From a copyright perspective, an AI industry that meets these standards is one that:
1. values and respects the rights of creators and copyright owners and the importance of the copyrighted works they create;
2. does not make those rights and works subservient to the interests of AI developers; and
3. recognizes and respects the value of the already well-established and expanding licensing market for AI-related uses.
Many of our DMLA members are already using or plan to use generative artificial intelligence (GAI) to aid in the creation of a wide range of works that benefit society, and some are themselves developers of GAI technologies. As users and developers of GAI technologies, as well as owners of copyrighted works that are used to train GAI models and are sourced using retrieval-augmented generation (RAG)3 , DMLA members are well-positioned and have a strong desire to see both the AI and creative industries thrive and continue to dominate globally.
Moreover, as long-time licensors and licensees of copyrighted works for more traditional uses, DMLA members recognize the value of and have embraced a free market for licensing their works for GAI training and for RAG. GAI licensing should not be viewed as an obstacle to the development and progress of GAI technologies. Frankly, the real hindrance to AI innovation would be a system in which copyright licensing is undermined, incentives for human creation and scientific research are diminished, and future AI models are trained on an ever-increasing amount of synthetic material, rather than robust, innovative, creative, and reliable original human works.
Copyright is not a barrier to GAI development. Copyright serves as the underlying foundation that ensures trustworthy and beneficial GAI development. We submit these comments to ensure that any policies adopted by the Administration: (i) are developed and implemented with a respect for and recognition of longstanding copyright laws and policies that make America the global leader in the creative and digital industries and in AI, and (ii) do not hinder the development, deployment, 3 For a discussion on retrieval-augmented generation technology and the copyright infringement implications, see U.S. Copyright Office, Copyright and Artificial Intelligence, Part 3: Generative AI Training, at 30 (2025), https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-3-Generative-AI-Training-Report-Pre- Publication-Version.pdf. and adoption of GAI technologies within the United States by disrupting the existing and growing copyright licensing market for AI-related uses.
Established Copyright Laws are Rooted in the Constitution and Promote Human Flourishing Alongside Economic Competitiveness
The RFI asks for input from various stakeholders on priorities for regulatory reform or other agency action necessary to promote AI innovation and adoption. As we discuss in more detail below, the successful development, deployment, and adoption of GAI technologies within the United States is consistent with and dependent on the goals of America’s copyright system, enshrined in Article I, Section 8, Clause 8 of the Constitution. Known as the “IP Clause,” it grants Congress the power “to promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries.” IP laws, including copyright laws, are what enable human authors to earn a living by creating and innovating, which is the key to securing American sustained economic competitiveness and global leadership in both its creative and AI industries.
It is essential that any report or policies resulting from the RFI be consistent with the Constitutional guarantees that protect copyright owners and human creators without whom GAI systems would not and cannot continue to exist. Additionally, any policies emanating from the RFI must consider and be consistent with President Trump’s Executive Order on AI, which makes clear that AI policies must “promote human flourishing, economic competitiveness, and national security.”4 (emphasis added). Further, the Administration’s AI Action Plan makes clear that “[h]igh-quality 4 Exec. Order No. 14179, 3 C.F.R. (2025). Data has become a national strategic asset.”5 There is no question that human-created copyrighted works produced by America’s creative industries have become the backbone of the high-quality “data” AI companies want to use to train or augment generative models, and they are unequivocally a national strategic asset that should be protected with the same vehemence that patented AI chip technology and other forms of AI-related intellectual property are protected.
Existing U.S. copyright laws, as detailed further below, are carefully balanced to provide essential protections along with important flexibilities, a deliberate framework that must not be altered for GAI. From broadcast content, film and TV shows, and journalism to sound recordings, works of visual arts, books, and everything in between, the ingestion of copyright-protected works for GAI training and augmentation is one of the central controversies related to the development of GAI technologies. Whether the unauthorized ingestion of copyright protected works for training constitutes copyright infringement or whether it qualifies for U.S. copyright law’s four-factor fair use exception is an issue that has become the focus of over fifty ongoing federal lawsuits, and it’s one that will and should continue to be decided by the courts on a case-by-case, fact-intensive basis.
Federal courts have been applying fair use principles for over a century, over the course of various technological advancements like the photocopy machine, the VCR, the Internet, digital music services, and many other new technologies. Courts are capable of applying the fair use doctrine to novel questions surrounding disruptive technologies, and they are best positioned to do so with GAI. There is no need currently to change copyright law or create a GAI exception in the law. This is not just the view of a broad consensus of the copyright industries, it is also the view of numerous AI companies, and the diverse industry groups that represent them.
There are many areas related to AI where the Administration may feel the need to take action to help facilitate U.S. world dominance, but copyright is not one of those areas. Any attempt to alter long-standing and balanced copyright laws would have the effect of destroying the incentive for humans to create and ultimately be harmful to U.S. dominance in AI.
Copyright Contributes to America’s Economic Growth and Job Creation
While AI is a significant and welcome contributor to the economy, the contributions of U.S. creative industries—made possible through copyright law—have been one of the most significant contributors to the U.S. economy and to job creation for decades. A recent report on the economic impact of copyright by the International Intellectual Property Alliance notes that, in 2023, the core copyright industries contributed more than $2 trillion to the U.S. gross domestic product (GDP) (accounting for 7.66% of the U.S. economy) and employed 11.6 million workers (or 5.43% of the
workforce). In addition to growing at a rate more than three times that of the rest of the economy, the report notes that the core copyright industries:
(1) make up an increasingly large percentage of value added to GDP;
(2) create more and better paying jobs than other sectors of the U.S. economy;
(3) grow faster than the rest of the U.S. economy;
(4) contribute substantially to U.S. foreign sales and exports, outpacing many industry sectors;
and
(5) make significantly large contributions to what the [U.S. Bureau of Economic Analysis] defines as the digital economy, which does not even encompass the full scope of the copyright industries’ digital activities.
Copyright industries are an invaluable asset to the U.S. economy because the exclusive intellectual property rights afforded by copyright incentivize investment in the creation and dissemination of new expressive works and allow copyright owners to recoup that investment. The U.S. continues to be the world leader in IP—an attribute that contributes significantly to this country’s vast cultural influence, leading role in scientific and medical advancements, and standing as the world’s leading economy. U.S. leadership in IP is an integral part of sustaining dominance in AI, and any policies that result from this RFI must consider the effect any actions or policies may have on copyright’s contributions to the economy and job creation.
Free Market Copyright Licensing Enables AI Development
Promoting private sector free markets and a robust voluntary licensing ecosystem is essential to ensuring American dominance in AI. Copyright law enables creators and copyright owners to supply AI companies with flexible and responsive solutions through tailored licensing and business models for GAI development. The ability of creators and copyright owners to enforce their rights in the works they create is crucial because it incentivizes the further creation and proliferation of high- quality creative works which form the basis for GAI development. Without copyrighted works to train and augment GAI models, the models cannot generate high-quality outputs. The growing number of licensing and partnership deals between AI companies and rights holders being reached with each passing day demonstrates these points.
Since the rise of GAI technologies a few years ago, the number of private sector free-market licensing agreements between copyright owners and AI companies has grown significantly. Increasing numbers of copyright owners, particularly news, magazine, scientific and academic publishers and image/media companies are licensing their copyrighted works to AI companies for commercial uses and have been doing so for many years.10 This shows that the market is working and there does not need to be any change in copyright law or any new policies that could disrupt that market.
While the GAI-copyright licensing market has grown over time, this growth will be destroyed by changes to copyright law that create exceptions for GAI.11 Nobody disputes that GAI companies and developers must pay for and invest in computer chips and cloud infrastructure. It is part of the cost of doing business in a free market. So, too, is free-market licensing of copyrighted works. To think otherwise would be detrimental to American economic competitiveness, in light of the fact that strong copyright laws can and already have been shown to foster GAI innovation as it forms the basis of competitive GAI products, not to mention copyrighted works’ own, direct benefit to the American economy and balance of trade with foreign nations.
For a list of the dozens of licensing agreements struck between copyright owners and AI companies, see https://copyrightalliance.org/artificial-intelligence-copyright/licensing/. 10 In the U.S., just a few public examples of recent licensing solutions, initiatives, partnerships, and agreements for AI use of copyrighted works include those launched from or created by Authors Guild, Created by Humans, Dataset Providers Alliance, Copyright Clearance Center, Elsevier, Getty Images, Shutterstock, Jstor, Sage Journals, Rightsify, Universal Music Group, and other major media publishers including the Associated Press, Axios, Condé Nast, News Corp, The Atlantic, Vox Media, Dotdash Meredith, Fortune, Time, Entrepreneur, The Texas Tribune, and WordPress.com.
Proposals to change the existing legal framework will undermine the market for responsible GAI collaborations by creating a strong incentive for GAI developers to wait for a new legal environment where working in good faith with copyright owners is not necessary or beneficial for their bottom line.
No policy should be adopted in response to GAI that interferes with the free market and the freedom of copyright owners and AI companies and developers to enter into licensing agreements. The marketplace should continue to properly value and incentivize creativity, and any AI policies
adopted by the U.S. Government should not interfere with the right of copyright owners to choose whether and how to license their works for GAI purposes. Copyrighted works provide immense value to GAI developers, and they can and should pay for that value—as many are already doing today. In other words, copyright law sets the conditions for the market to prevail and for the U.S. to maintain its position as a global leader in both the AI and creative industries. Any policy that disrupts these existing licensing markets or adversely impacts the incentives to license in the future would not only harm America’s vibrant creative economy but would also hinder the development, deployment, and adoption of GAI technologies within the United States. If GAI licensing markets are overridden by new laws or regulations, the result would be to reduce the incentives for humans to create. And if humans are no longer creating copyrighted works for GAI models to train on, these models have no other alternative than to train on synthetically generated AI outputs. A future where GAI systems train on datasets comprised mostly or entirely of AI generated works would result in poor quality, biased, and incorrect outputs, ultimately leading to model collapse.
It is essential that we never get to the point of model collapse. The best way to prevent model collapse is to ensure that the incentives to create new creative works spurred by U.S. copyright law remain strong and vibrant and that the licensing markets for copyrighted works are supported and encouraged, not thwarted, by the Administration, Congress, and the courts.
Copyright Transparency Leads to AI Dominance
Developers of publicly available GAI models that ingest the copyrighted works of third parties without a license should be required to satisfy transparency standards related to the collection, retention, and disclosure of the copyrighted works they use to train GAI models. Adequate transparency regarding ingestion of unlicensed copyrighted works is vital to ensuring that copyright owners’ rights are respected alongside the advancement of GAI technologies, and it should not be considered an impediment to the development and dominance of the U.S. AI industry.
Best practices from corporations, research institutions, governments, and other organizations that encourage transparency around GAI ingestion already exist that enable users of GAI systems or those affected by their outputs to know the provenance of those outputs.13 There is no reason these same responsibilities should not also apply to GAI ingestion of unlicensed copyrighted works. It is vital that GAI developers be required to maintain adequate and proportionate records of copyrighted works they neither own nor license that were used to train and to make those records publicly accessible and searchable as appropriate. Adequate and appropriately scoped transparency and record-keeping requirements benefit copyright owners by enabling them to learn whether and how their works have been used to train GAI models, and benefit GAI developers in that transparency promotes consumer trust.14 Consequently, transparency by businesses that offer GAI systems to the public is a crucial component of any AI policy.
The National Institute of Standards and Technology recently explained that “[e]xternal parties’ willingness to use or build on datasets and models depends partly on the extent to which they feel equipped to understand, in a manner that Proposed Copyright Exceptions and Opt-Out Systems Hinder AI Innovation.
AI innovation depends on quality GAI training, and quality GAI training depends on valuable training material in the form of copyrighted works created and protected under the framework of established copyright laws. Proposals calling for blanket copyright exceptions in the context of GAI, including those made by foreign governments, should be identified as barriers to American AI innovation and be rejected. These exceptions and proposals would weaken copyright protections while exacerbating longstanding copyright piracy issues, hindering the sustainability and competitiveness of America’s creative sector and its ability to contribute to U.S. economic growth, job creation, and development of quality materials and data for GAI training. We have already seen challenges to IP protection arise in the context of GAI technologies being developed in China, which steals U.S. creative and cultural output to gain an advantage. America’s IP laws, including our robust protections for our creators and innovators and our strong stance against piracy, is what sets us apart from China and other countries that unfairly circumvent or weaken copyright owners’ rights. Strong IP and copyright protections are ultimately what give the U.S. an advantage over those countries, and if we neglect those principles our advantage will be lost.
Importantly, U.S. copyright law is unequivocally an “opt-in” regime and the driving force behind the already dynamic GAI licensing market for copyrighted works. There are many small-to- medium-sized AI companies that have built their entire businesses upon commitments, partnerships, and licenses secured with copyright owners under the copyright legal framework. suits their needs, how well-considered the processes and decisions behind the product are and how the product could affect various interested parties.” Extended Outline: Proposed Zero Draft for a Standard on Documentation of AI Datasets and AI Models. National Institute of Standards and Technology, U.S. Department of Commerce, Sept. 2025,
Opt-in, permissions-based agreements and licenses have resulted in more, not fewer, partnerships between the AI and creative sectors. On the other hand, “Opt-out” systems that permit AI companies to use works unless the copyright owner affirmatively objects will act as a hindrance to AI innovation. We strongly oppose any enactment of an opt-out regime as it would act as a barrier for rightsholders and AI companies alike, particularly in the GAI licensing marketplace. GAI opt-out schemes don’t work. Those who propose them fail to consider the practical difficulties of their implementation. For example:
(1) the practical effects of opt-out, particularly regarding works already used to train GAI, are negligible given that it is challenging to remove entire works at scale from a GAI model; and (2) despite a copyright owner opting out, copies of works may still be included in GAI datasets through other downstream means, such as when copies are scraped from a licensee of the copyright owner, a third-party platform (that fails to recognize an opt-out), or a piracy site where a copy has been posted without authorization. While some technical solutions may assist with opt-out, they typically have significant limitations because they are only effective to the extent the opt-out mechanism is recognized and respected,
and because these tools are often not designed to be targeted to address scraping for GAI ingestion.15 Comprehensive opt-out is also an unachievable goal as copyrighted works often exist in multiple places on the internet that make it nearly impossible for a copyright owner to apply the opt-out indicator to every copy of a work. For example, a single song can be streamed on a digital streaming platform, played as the background music of a user uploaded video on a social media platform or in advertisements, or displayed as notes or lyrics on a website. Though tools are being developed and improved upon, it is still impossible for the copyright owner to successfully opt out in a way where every single downstream use would be tagged with the proper recognized and respected opt-out signal to prevent GAI scraping and use. The ongoing discussions on this issue in the context of the EU AI Act clearly demonstrate that no workable opt-out mechanism currently exists or is likely to exist in the future.
Moreover, copies of works that are available on pirate sites are even further removed from the copyright owner’s control, and it is well-known that AI companies have used pirated copies of creative works to train their GAI models and have even proliferated pirated copies themselves during the GAI development process.17 The recent settlement in the Bartz v. Anthropic case involved Anthropic’s use of pirated works to train its GAI, and illustrates why AI companies should not source training materials from piracy websites in favor of legitimate sourcing through GAI licenses with rightsholders.18 Policies which give even a hint of legitimizing copyright piracy encourages a global race-to-the bottom, ultimately hurting American creative and AI businesses and impeding their abilities to create, among other things, valuable training materials and data to contribute to American AI innovation.
An opt-out regime fails to address or ameliorate any of these problems and certainly does not afford the copyright owner any semblance of control. For these same reasons, there is currently a high level of uncertainty over what constitutes an effective opt-out,1819 and this uncertainty is being exploited by some GAI developers who continue to train on scraped content despite legitimate efforts from copyright owners to opt out. In sum, opt-out does not and will not work and proposals to enact such a system are a barrier to AI innovation and should be rejected.
Conclusion
DMLA strongly urges the Administration to recognize that existing copyright laws support, rather than hinder, AI innovation. The current legal and licensing structure provides a balanced foundation for protecting creative rights while advancing technology. Any regulatory reforms that weaken copyright protection or introduce opt-out systems would harm both the creative economy and AI competitiveness. Instead, any policies should promote transparency, support responsible innovation, and uphold intellectual property as a cornerstone of America’s leadership in AI and creative industries.
We appreciate the opportunity to submit these comments, and we are happy to answer any additional questions.
Respectfully submitted,
Joe G. Naylor
President
Digital Media Licensing Association
c/o Cowan, DeBaets, Abrahams & Sheppard LLP
60 Broad Street, 30th Floor
New York, New York 10004
October 27, 2025