New USPTO Guidance Hones AI-Related Patent Eligibility Criteria

The U.S. Patent and Trademark Office has announced new guidance on patent subject matter eligibility as it relates to AI inventions. The guidance features three new subject matter eligibility examples (Examples 47, 48, and 49). These examples are part of the ongoing effort to clarify how inventions involving artificial intelligence (AI) are evaluated under the U.S. patent system. The guidance aims to help examiners and practitioners better understand how to apply the criteria for patent eligibility, particularly in the context of rapidly evolving technologies like AI.

In recent years, AI has become a pivotal technology across various industries, driving innovation in fields such as health care, cybersecurity, and data analysis. However, the unique nature of AI, often involving complex algorithms and data processing techniques, has raised challenging questions about what constitutes patentable subject matter. The USPTO’s examples serve as hypothetical scenarios to illustrate how the existing legal framework, particularly under Section 101 of the U.S. Patent Act, is applied to AI-related inventions.

These new examples are intended to complement the existing guidance on subject matter eligibility, providing concrete illustrations of how claims are analyzed for patent eligibility. They highlight critical factors such as the integration of abstract ideas into practical applications and the necessity of demonstrating a technical improvement or inventive concept. Through these examples, the USPTO aims to ensure that the patent system continues to encourage innovation while maintaining a clear and consistent standard for what constitutes patentable subject matter.

Example 47: Anomaly Detection

This example involves claims related to the use of artificial neural networks (ANNs) to identify or detect anomalies, such as in network security. Claims that integrate the abstract idea into a practical application, such as improving network security, are considered eligible, while those that do not are deemed ineligible.

Example 48: Speech Separation

This example discusses claims related to using deep neural networks (DNNs) to separate speech signals from different sources. The eligibility of these claims depends on whether they provide a practical application, such as improving speech-to-text transcription, or merely cover abstract ideas without significant practical integration.

Example 49: Fibrosis Treatment

This example addresses claims related to the use of AI to analyze medical data for fibrosis treatment. The eligibility hinges on whether the AI implementation provides a concrete improvement in medical treatment processes or remains abstract.

These new examples provide valuable insights into the patent eligibility of AI-related inventions and illustrate the importance of integrating abstract ideas into practical applications. As AI continues to play a transformative role in various industries, understanding how to navigate the complexities of patent eligibility is crucial for protecting innovative solutions. Practitioners are encouraged to thoroughly review this latest guidance and consider its implications in their practice. By aligning their patent strategies with the principles outlined in these examples, practitioners can better ensure the successful prosecution of AI-related patents and contribute to the continued growth and innovation in this dynamic field.

Ballard Spahr’s Artificial Intelligence Team leverages our lawyers’ experience with AI issues specific to a range of sectors, including life sciences, health care, financial services, retail and manufacturing, e-commerce, and technology. The Team includes members of the firm’s Patents Group, internationally known for handling patent prosecution, portfolio development and management, and collaboration in transactions for innovative businesses, including well-known global brands. 

Copyright © 2024 by Ballard Spahr LLP.
www.ballardspahr.com
(No claim to original U.S. government material.)

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, including electronic, mechanical, photocopying, recording, or otherwise, without prior written permission of the author and publisher.

This alert is a periodic publication of Ballard Spahr LLP and is intended to notify recipients of new developments in the law. It should not be construed as legal advice or legal opinion on any specific facts or circumstances. The contents are intended for general informational purposes only, and you are urged to consult your own attorney concerning your situation and specific legal questions you have.