Publication

Nov 1, 2023

White House Directs Copyright Office and USPTO to Provide Guidance on AI-Related Issues

Click for PDF

As our colleagues reported in this Seyfarth Shaw Legal Update, President Biden signed a comprehensive Executive Order addressing AI regulation across a wide range of industries and issues. Intellectual property is a key focus. The Order calls on the U.S. Copyright Office and U.S. Patent and Trademark Office to provide guidance on IP risks and related regulation to address emerging issues related to AI.

Patent Law

President Biden’s directive instructs the Under Secretary of Commerce for Intellectual Property and Director of the U.S. Patent and Trademark Office (USPTO), Kathi Vidal, to provide guidance on AI’s role in patent inventorship and support for patent practitioners. Specifically:

  • AI Inventorship: The Executive Order mandates the USPTO Director to publish, by the end of February 2024, guidance to the USPTO’s patent examiners and applicants addressing inventorship and the utilization of AI, including generative AI, in the inventive process. This comes amidst a backdrop where AI models currently cannot be recognized as inventors, a stance reaffirmed in the Federal Circuit decision Thaler v. Vidal. See our prior post on issues raised by Thaler.
  • Additional Guidance: Subsequent to the initial guidelines, the Executive Order requires the USPTO Director, by July 2024, to issue further guidance on other considerations “at the intersection of AI and IP.” The Executive Order further states that those other considerations might encompass support mechanisms for patent practitioners and updated guidelines on patent eligibility concerning AI innovations, if the USPTO Director “deems necessary.” It appears this guidance may shed further light on various patent eligibility issues under 35 U.S.C. § 101. See our prior post discussing some of the patentability issues for AI.
  • Support for Patent Practitioners: It’s unclear whether the guidance under AI Inventorship or the Additional Guidance will address the use of AI in the patent application drafting process.

Copyright Law

The directive also instructs the Under Secretary and Director Vidal to consult with the Director of the Copyright Office to issue guidance by July 2024, or 180 days after the Copyright Office’s forthcoming study on AI and copyright issues. These recommendations are to focus on potential executive actions related to copyright and AI. They should also address issues discussed in the Copyright Office’s study, including the scope of protection for works produced using AI and the treatment of copyrighted works in AI training. Although the Executive Order does not provide great detail, we expect guidance on:

  • AI-Generated Art: Copyright issues arise when AI systems generate art, music, or literature. If an AI creates a masterpiece, who is the rightful owner of the copyright? Is it the programmer, the entity that deployed the AI, or the AI itself? We have previously written about this issue here and here.
  • Fair Use: The Executive Order encourages data sharing and research. This might conflict with copyright laws if the data shared includes copyrighted material. Fair use and licensing agreements become significant concerns, as AI developers utilize existing datasets and copyrighted content for training and testing. We have written about lawsuits surrounding these issues here and here.
  • Commercial Use: AI technologies are widely used for content generation, such as automated news articles or product descriptions. How will copyright affect the commercial use of AI-generated content? Should AI-generated content be considered as intellectual property, and if so, who should benefit from it?
  • Orphan Works: AI has the potential to identify and restore orphaned works – copyrighted works whose owners are unknown or untraceable. This raises questions about the ownership and fair use of AI-discovered orphan works.

AI-Related IP Risk Mitigation Program

The Executive Order also contains provisions relating to AI and IP risk mitigation. In the Order, President Biden has directed the Secretary of Homeland Security, in consultation with the Attorney General, to develop a training, analysis, and evaluation program by the end of April 2024. This risk mitigation program comes against the backdrop of President Biden’s remarks on America’s need to “continue to lead on AI.” In the Department of Homeland Security’s press release regarding President Biden’s Executive Order, DHS observed, “Protecting AI intellectual property (IP) is critical to U.S. global competitiveness. IP theft threatens U.S. businesses, impacts American jobs, and negatively [a]ffects our national security.”

Many provisions of the Executive Order prioritize security concerns regarding large, cutting-edge AI models. For example, when American companies train the largest cutting-edge “foundation models,” the Executive Order requires companies to disclose “the physical and cybersecurity protections taken to assure the integrity of that training process against sophisticated threats” and also “the physical and cybersecurity measures taken to protect … model weights.” Model weights are key in the context of intellectual property disputes, because they affect how much influence input (often copyright-protected) will have on output (the subject of many ongoing lawsuits, as noted above). In an earlier announcement over the summer, the Biden Administration has called model weights “the most essential part of an AI system.”

Ultimately, the scope of “AI-related IP risks” addressed by the risk mitigation program mandated by the Executive Order extends beyond the realm of the largest foundation models, encompassing a broader range of AI-related IP challenges. The program will include dedicated personnel for collecting and analyzing reports of AI-related IP theft in a more general context. The program will also promote broad collaboration and information sharing among various federal, state, and local government agencies, including the FBI and U.S. Customs and Border Protection, as well as international organizations, with the goal of creating a more unified front against AI-related IP theft. The Executive Order also directs the development of guidance and resources for private sector actors to mitigate AI-related IP theft.

As part of the risk mitigation program, DHS will share information and best practices with AI developers and law enforcement personnel to identify incidents, inform stakeholders of legal requirements, and evaluate AI systems for IP law violations.  This last point, about evaluating AI systems for IP law violations, suggests that DHS and DOJ will also be focused on the ways in which AI systems (and the training data used by those systems) might themselves violate IP laws—a question that has already given rise to extensive litigation (see above).

The program envisioned in the Executive Order will also assist the Intellectual Property Enforcement Coordinator in updating the Joint Strategic Plan on Intellectual Property Enforcement to address AI-related issues.

On the Horizon

IP owners and those using others’ inventions and content in their business will need to keep a close watch on USPTO announcements regarding the forthcoming guidelines, which could redefine AI’s involvement in patent inventorship and provide support for patent practitioners. Inventors should also evaluate existing patent drafting processes and envisage how the integration of AI, in light of forthcoming USPTO guidelines, could augment efficiency and compliance.

On the copyright side, we may see legislative reform as Congress determines whether it needs to adapt the Copyright Act to the AI era. Defining AI as a creator, establishing new copyright ownership criteria, and creating a fair licensing framework are possible steps. Clearer guidelines on fair use of copyrighted material in AI development may also be necessary to strike a balance between encouraging innovation and respecting intellectual property rights. Meanwhile, AI developers, content creators, and copyright holders may need to negotiate licensing agreements to specify how AI-generated content can be used and who benefits from it.

We will continue to monitor these evolving issues and provide updates as we receive additional agency guidance pursuant to the Executive Order.  Be on the lookout for continued updates from us and from our colleagues across the firm as Seyfarth provides additional insight and analysis on this and other aspects of the Executive Order in the weeks and months to come. 

You can hear more about the Executive Order and other AI issues during Seyfarth’s upcoming webinar, hosted by Lexology, on November 15, 2023.