Written by: Shirin Tehrani
The United States Patent and Trademark Office (USPTO) recently released updated guidelines designed to clarify certain points within the USPTO’s patent subject matter eligibility for Artificial Intelligence (AI). Effective from July 17, 2024, the revised guidelines provide updated examples and discuss recent Federal Court decisions on patent subject matter eligibility.
Federal court decisions on patent subject matter eligibility have been shaped by Supreme Court rulings, such as Bilski, Alice, and Mayo. These decisions established a two-step test to determine patent eligibility. The first step assesses whether the invention falls into a statutory category under 35 U.S.C. 101, specifically whether it claims a process, machine, manufacture, or composition of matter. The second step involves a two-prong framework: first, determining if the claim recites a judicial exception (an abstract idea, law of nature, or natural phenomenon). If it does, the invention is ineligible. If not, the second prong assesses whether the claims amount to “significantly more” than the judicial exception. If it does, the claim may be eligible for patent protection.
The updated guidelines offer a detailed discussion on evaluating whether a claim recites an abstract idea, such as mathematical concepts, certain methods of organizing human activity, or mental processes. The updated guidelines also provide information on how to assess the “significantly more” standard of the Alice/Mayo test.
Examples 47, 48, and 49 of the updated guidelines illustrate how the USPTO applies these principles:
- Example 47: Discusses the use of ASICs for AI-based anomaly detection, emphasizing the patent eligibility of particular hardware designs that improve network security.
- Example 48: Discusses AI methods for speech separation indicating eligibility when claims specify technical improvements like processing and parsing the speech signals.
- Example 49: Discusses how AI models in personalized medicine may be patent eligible when the claims include detailed and specific treatment methods to transform abstract ideas into practical solutions.
While the updated guidelines are now in effect, the USPTO has invited the public to share their experiences via regulations.gov. The feedback will inform future revisions and interpretations to help the guidelines evolve with AI technology. Patent attorneys and industry stakeholders are encouraged to engage with these examples to better navigate patent application process and adhere to evolving legal standards. However, challenges like restrictive categorization, potential discouragement of certain AI innovations, regulatory uncertainties, and difficulties in aligning guidelines with technological progress must be addressed to foster a robust and competitive AI landscape in the United States.
The USPTO’s updated guidelines represent a proactive step towards coordinating patent eligibility assessments for AI innovations. By emphasizing technical solutions and practical integrations, the guidelines aim to foster innovation while upholding rigorous standards of patentability. As AI technologies continue to evolve, understanding and applying these guidelines will be fundamental for securing intellectual property protection and driving future advancements in the field.
Despite the challenges in defining patent eligibility for AI, the USPTO’s guidelines offer a detailed framework that balances the need for innovation with legal clarity, promoting a dynamic and competitive AI environment in the United States.