Greg Derin
In June in Loper Bright Enterprises v. Raimondo, the U.S. Supreme Court overruled Chevron U.S.A. Inc. v. National Resources Defense Council Inc.
Chevron required federal courts to defer to reasonable statutory construction by agencies to which Congress had delegated authority where the statutes were deemed ambiguous, and the agency's construction was permissible.
The Chevron doctrine was deemed by many to have created consistency in statutory construction and required courts to defer to the presumed expertise of administrative agencies.
In the aftermath of Loper Bright, extensive commentary has focused on the legal propriety of the decision, as well as the motivations for and implications of the ruling. Many argue that Chevron has been effectively dead for more than a decade, diminished by judicially created exceptions.
Where comment has dealt with the anticipated impact on regulatory oversight, it has focused on areas such as public health, the financial sector and environmental protection. As one who mediates intellectual property disputes, I have pondered the possible use of Loper Bright by those who would intrude on the historic functions of the U.S. Copyright Office and the U.S. Patent and Trademark Office.
A battleground that has surfaced recently involves artificial intelligence. As the presidential election heats up, a number of Silicon Valley entrepreneurs have lined up behind the Republican nominee, Donald Trump.
The Washington Post reported in June that a group of the candidate's allies are drafting an AI executive order, which would launch a series of "Manhattan Projects" to develop military technology and review unnecessary and burdensome regulations. Their goal is to accelerate the development and deployment of AI tools and create industry-led agencies to evaluate AI models and security systems.
This agenda is in direct contrast to President Joe Biden's agenda, as expressed in a promulgated executive order, which seeks to regulate the development of AI to assure its safe use and deployment and the security of privacy rights of individual citizens.
These potentially conflicting agendas have national and global consequences and can define who controls privacy and security for individual Americans — the government or entrepreneurs. In the narrower sphere in which many litigate, how might the demise of Chevron affect developing copyright law relating to artificially created content?
For more than a year, the Copyright Office has been soliciting comments and enforcing its views on the quantum of human authored content required of submitted works before they qualify for copyright registration. With Kris Kashtanova's "Zarya of the Dawn," the Copyright Office concluded in February 2023 that human-authored text — combined in a compilation with images generated by an AI platform — constituted a copyrightable work, but the individual images generated by the AI platform were not themselves copyrightable.
A month later, on March 16, 2023, the Copyright Office published its copyright registration guidance on works containing material generated by AI. In the guidance, the office iterated its long-standing view that "copyright can protect only material that is the product of human creativity. Most fundamentally, the term 'author,' which is used in both the Constitution and the Copyright Act, excludes non-humans."
Recognizing the development of AI, the Copyright Office noted that it would conduct case-by-case reviews to determine how AI tools were used to contribute to a work submitted for registration. The guidance then provided advice on submitting an application for works containing AI-generated material.
The position taken in the guidance referenced extant statutory authority and case law. Both before and after the issuance of the guidance, courts have ruled in a manner consistent with the position articulated by the Copyright Office.
In the fall of 2023, the Copyright Office conducted a study and invited comment regarding further regulations applicable to AI-generated material, demonstrating its ongoing commitment to monitoring the evolving technology and practitioners' views and its legal implications.
While the Copyright Office and courts have thus far been aligned on the historic requirement of human contribution to the creation of copyrightable material, AI has introduced not only new, but quickly evolving and subtle questions. The Copyright Office has addressed these by noting the requirement of case-by-case analysis to distinguish purely mechanical contributions from creative ones.
AI will continue to evolve and may do so without existing guardrails. With the demise of Chevron, industry leaders may seek to shift the balance of power to courts to exercise more independent statutory interpretation without constraints from the Copyright Office.
It is not clear that the role of the Copyright Office will fall within the ambit of ambiguous delegated authority. However, for anyone wishing to gain more control of the AI marketplace and bridling at restraints created by the guidelines or its later iterations, it is natural to assume that Loper Bright is a prelude to challenges.
A shift to more judicial independence is likely to manifest itself in analyses concerning fair use, deference given to the office's registration determinations, and regulations covering an array of traditional agency interpretative areas such as Digital Millennium Copyright Act exemptions and the scope of safe harbors and exceptions.
Thaler v. Perlmutter in the U.S. District Court for the District of Columbia last year provides an example as to how the process has worked historically, and how it might be affected going forward. Stephen Thaler claimed that his computer system generated a work of visual art "of its own accord," listing the computer system as the author,
Thaler sought registration from the Copyright Office, with the copyright thereafter to be transferred to him as owner of the generating computer system. The Copyright Office denied his application due to lack of human authorship; as no human created the work, there was no valid copyright to transfer.
On summary judgment, the district court sided with the Copyright Office. While adaptable to changing technologies, the court concluded that "human creativity is the sine qua non at the core of copyrightability, even as that human creativity is channeled through new tools or into new media."
The Thaler court engaged in a thoughtful analysis consistent with interpretations and rulemaking of the Copyright Office, the U.S. Constitution and prior case law. But technology will continue to evolve rapidly. The Copyright Office has devoted significant resources, attention and expertise to analyzing and soliciting comments from the copyright community to consider nuances of AI input and output.[1]
Will courts have the time, expertise and independence to do so in the future? Will all of the federal circuits do so consistently? Any practitioner in the area who has tried to find intellectual consistency and guidance in the Supreme Court's recent copyright decisions — including Andy Warhol Foundation for the Visual Arts Inc. v. Lynn Goldsmith last year — will wonder.
Not all agencies will have their mandate curtailed following Loper Bright. But it is a certainty that those who wish to control a marketplace and feel that their chances are better in court than strictly at an administrative level, will invest in such a challenge. Copyright issues are difficult enough without bright lines dimming beyond recognition.
On July 16, the USPTO issued new guidance on patent eligibility for inventions involving AI. Already encumbered by conflicts between USPTO determinations and interpretations by the courts, the slippery slope created by Loper Bright may soon become unmanageable for all intellectual property involving AI.
Greg Derin is a mediator and arbitrator at Signature Resolution.
The opinions expressed are those of the author(s) and do not necessarily reflect the views of their employer, its clients, or Portfolio Media Inc., or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.
[1] See also, Munro v. United States Copyright Office
, 2024 WL 1156519 (USDC D.C. March 18, 2024)(on appeal)(demonstrating the significance of the Office's determinations in registration matters).For a reprint of this article, please contact reprints@law360.com.
