This article originally appeared as a guest column in the Journal Record on Aug 17, 2023.
Cody J. Cooper
Anyone paying attention to modern culture is aware of the recent groundbreaking advancements in artificial intelligence. With tools such as ChatGPT, one simply needs to enter a query into an interface, and within seconds the software seems to magically generate content that would take most humans hours to compose.
AI is able to generate “expressive material,” defined by the U.S. Copyright Office as “AI output that, if it had been created by a human, would fall within subject matter of copyright as defined in section 102 of the Copyright Act.” The USCO maintains that AI-generated output will not be classified as creative works because “[i]f a work’s traditional elements of authorship were produced by a machine, the work lacks human authorship, and the Office will not register it.” Even though AI output does not currently qualify for copyright protection (an area of continued discussion), liability still lurks.
Since AI is “trained” on vast quantities of preexisting human-authored works, use of some content and production of output may constitute copyright infringement. AI research company OpenAI argues that the use of copyrighted works to train AI programs should be considered fair use under 17 U.S.C. § 107, claiming that the AI-generated work is transformative. Although USCO describes “transformative” as “uses that add something new, with a further purpose or different character, and do not substitute for the original use of the work,” and is “more likely to be considered fair,” OpenAI’s argument is unsettled.
Ultimately, AI copyright infringement and fair use will be viewed like any other infringement. The most recent Supreme Court case provides analysis of fair use in the context of copyright infringement. In essence, the May 18 decision in Andy Warhol Foundation for the Visual Arts v. Goldsmith held that Warhol’s silk-screen series of portraits of pop-star Prince violates the copyright of photographer and author of the original image because it was not sufficiently transformative. This analysis will apply for AI-generated content, and courts will essentially review whether: (1) AI had access to the copyrighted material and (2) the AI-generated output is substantially similar to the copyrighted work. Who then remains ultimately responsible for the infringement is another unsettled question.
Ultimately, questions regarding AI-generated output and its protection and potential legal liabilities are much like AI technology, constantly evolving.
About the author:
Cody J. Cooper is a Director at the law firm of Phillips Murrah. Cody is an experienced litigator and a licensed patent attorney who represents individuals and companies in a wide range of business litigation and intellectual property matters. His practice primarily concentrates on complex, commercial litigation in state and federal courts as well as intellectual property, including litigation and patent, trademark and copyright prosecution, and cybersecurity and privacy matters.
CONTACT: cjcooper@phillipsmurrah.com | 405.552.2405
Follow Phillips Murrah on social media:
