Washington, D.C. — TechNet, the national, bipartisan network of technology CEOs and senior executives, joined the Computer & Communications Industry Association (CCIA) and NetChoice in filing an amicus brief with the U.S. District Court for the Southern District of New York in the case of The New York Times v. Microsoft and OpenAI. The brief supports OpenAI’s argument that its generative AI model training practices comply with the longstanding copyright principle of fair use. Stated TechNet President and CEO Linda Moore:
“AI may be a new technology, but this is not a new issue. For over a century, U.S. copyright law has balanced the interests of creators with the need for creativity and innovation. Fair use has long protected the ability to learn from published works to create novel ideas and produce new outcomes, and that principle should apply here. Copyright law protects an author’s expression, not the underlying facts, ideas, or knowledge. To change that principle only when AI is involved would overturn over a century of American law and ensure basic facts are available only to a select few.
“It is essential that the court uphold America’s carefully balanced copyright law framework and ensure it continues to serve its intended purpose: protecting creators while fostering innovation and ingenuity.”
Specifically, the amicus brief argues:
- Fair use is essential to innovation. Copyright law strikes a balance between rewarding creators and allowing the public to access information, develop new technologies, and build on earlier work.
- Training generative AI models is a “transformative use.” Unlike selling unauthorized copies of a book, training an AI model does not reproduce or distribute copyrighted works to a new audience. Instead, AI models analyze existing works to learn patterns and relationships that enable the models to perform a wide range of new tasks. The purpose and character of AI training is therefore fundamentally different from that of the original works.
- Plaintiffs’ market dilution and licensing theories would improperly expand copyright protection. In copyright cases involving generative AI, plaintiffs have argued that training AI models creates works that compete with or dilute the market for copyrighted works, even when they do not reproduce protected expression. However, copyright law generally does not protect creators from competition by new, independently produced expression. Treating these effects as sufficient to defeat fair use would give copyright owners control over new technologies and creative works that fall outside the scope of copyright protection.