AI Law and Policy Developments Likely Will Be Seen This Year
One of 2023’s more significant — and potentially disruptive — developments in business and culture was the arrival of a slew of generative artificial intelligence (AI) systems. At the beginning of 2023, ChatGPT quickly captured our attention with its ability to deliver cogent, polished answers to most questions. In late spring, Google published a beta version of its generative AI system, and Meta released several of its own AI systems later in the year. By the end of 2023, everyone seemed to be talking about the variety and complexity of tasks that AI could perform, from analyzing big data and creating visual art to performing scientific research and composing music — and much more.
The penetration of generative AI into everyday life in 2023 was so rapid and so extensive that law and policy are now in a dead sprint to catch up. The law is increasingly being challenged to adapt to the new factual situations that AI presents, and the government appears to be learning the promise of AI and the potential risks of its uncontrolled use.
Patent law is one area in which AI is challenging the status quo. In August 2022, the Court of Appeals for the Federal Circuit squarely rejected the proposition that AI could be a patent inventor, holding that the Patent Act explicitly permits only a human to be an inventor (Thaler v. Vidal, 43 F.4th 1207 (Fed. Cir. 2022)). The Supreme Court declined to take up the issue in April 2023.
While it thus appears settled for now that technology conceived entirely by AI cannot be patented, technology conceived by an inventor using AI as a development tool remains patentable in appropriate circumstances. With patent applications involving AI increasing rapidly, the U.S. Patent and Trademark Office recognized in 2023 that there is an enhanced need for new policies for patent eligibility involving AI. In February 2023, the USPTO requested comments concerning the role of AI in inventorship, and it conducted a series of public listening sessions with various stakeholders later in the year to solicit input on potential solutions.
On Oct. 30, 2023, the Biden Administration substantially accelerated the USPTO’s study of AI policy with the issuance of an executive order that directs the USPTO to publish new guidance within 120 days on inventorship and the use of AI. Under the order, the USPTO’s guidance is to include examples of how the involvement of AI tools in the inventive process should be analyzed in patent examination. Therefore, during the first quarter of 2024, important new guidance can be expected from the USPTO that should clarify for inventors and patent practitioners how technology involving AI will be evaluated.
Effects in second IP arena
Copyright is another area in which existing law appears to be evolving to meet new demands created by generative AI. In March 2023, the US Copyright Office issued a statement clarifying that it will not register works “when an AI technology receives solely a prompt from a human and produces complex written, visual, or musical works in response” because “the ‘traditional elements of authorship’ are determined and executed by the technology — not the human user” (88 Fed. Reg. 16190, 16192 (March 16, 2023)). The Copyright Office also launched a new AI initiative in early 2023 “to examine the copyright law and policy issues raised by [AI], including the scope of copyright in works generated using AI tools and the use of copyrighted materials in AI training.”
After conducting listening sessions with artists, creative industries, AI developers, and the legal community, the Copyright Office published a notice in the Federal Register on Aug. 30, 2023, requesting information and views relating to issues raised by the recent advances in generative AI (88 Fed. Reg. 59942 (Aug. 30, 2023)). The notice asked for comments on a host of issues, such as whether there are circumstances in which a human using a generative AI system should be considered the “author” of the system’s output; whether AI-generated outputs implicate the exclusive rights of authors in preexisting works; and what legal rights should apply to AI-generated material that uses a particular person’s name or likeness or that imitates the artistic style of a human creator. The notice also asked for input on what rights, if any, should apply to AI-generated material if it were deemed ineligible for protection under copyright law.
As with the USPTO’s analysis of AI-related patent issues, the Copyright Office’s study of AI policy also was substantially accelerated by the executive order issued by the Biden administration on Oct. 30, 2023. The order directs the Copyright Office to issue new guidance on AI-related issues relating to copyright law by mid-2024. The order does not specify issues to be covered, but presumably, the Copyright Office’s forthcoming guidance will provide important new direction during 2024 on issues of authorship, fair use and rights of publicity relating to AI-generated outputs.
Issues before the courts
Questions about the intersection of copyright law and AI-generated material also surfaced in federal courts with increasing frequency during 2023. As generative AI moved into the spotlight, several lawsuits were filed last year in which copyright owners claimed that the use of their works as part of a dataset to “train” a generative AI system constituted copyright infringement. Several of those suits were filed by groups of authors challenging the use of their literary works to train generative AI systems (Silverman v. OpenAI Inc., No. 3:23-CV-03416-AMO; Kadrey v. Meta Platforms Inc., No. 3:23-CV-03417; Authors Guild v. OpenAI Inc., No. 1:23-CV-08292-SHS; Huckabee v. Meta Platforms Inc., No. 1:23-CV-09152-LGS; and Sancton v. OpenAI Inc., No. 1:23-CV-10211-SHS).
Another was filed by an owner of copyrights in photographic works which objected to a generative AI system’s use of its photos to train the system to generate similar visual works (Getty Images (US) Inc. v. Stability AI Inc., No. 1:23-CV-00135-GBW). Another action involves visual artists who objected to a generative AI system’s use of their copyrighted works to generate works in a particular artist’s style (Andersen v. Stability AI Ltd., No. 3:23-CV-00201-WHO).
While orders have been issued dismissing certain claims in some of the lawsuits, no final decisions have been entered to date, and some previously dismissed claims have been reasserted via amended pleadings. Some of the cases might produce major decisions of interest on AI-related copyright issues later in 2024.
The Biden administration’s executive order on AI tasks numerous parts of the federal government to study AI issues and to develop AI-related policies during 2024. For example, the order directs the Council of Economic Advisers to study the labor market effects of AI and the Department of Labor to evaluate potential AI-related displacements in the federal workforce. The Department of Justice is required to study the use of AI in the criminal justice system, and the Department of Health and Human Services is tasked with developing a strategic plan on the use of AI in research, drug and medical device safety, health care delivery, and public health. The order’s extensive reach is thus likely to shape policy on AI in 2024 and beyond relating to numerous aspects of business and government.
Generative AI is a potentially transformative technology that burst into ordinary American life in 2023 with an unexpected energy that existing law and policy were not fully prepared to handle. This year appears likely to be the year that law and policy begin a meaningful evolution that promotes its many potential benefits while containing its potential risks.
This article was first published by North Carolina Lawyers Weekly on Jan. 24, 2024 and is republished here with permission. ©2024 North Carolina Lawyers Weekly.