Panitch Schwarze | Demystifying Patent Eligibility Under 35 U.S.C. § 101 in the Age of Software and AI

Patent eligibility under 35 U.S.C. § 101 continues to be one of the most challenging aspects of U.S. patent prosecution, particularly for software- and AI-driven inventions. While the legal framework has long been shaped by judicial exceptions and evolving case law, recent developments at the USPTO and the Patent Trial and Appeal Board (PTAB) suggest a meaningful shift in how eligibility is being evaluated.

This article highlights key trends shaping § 101 analysis today and outlines practical considerations for innovators and patent applicants navigating this evolving landscape.

A Changing Approach at the USPTO

Recent changes in USPTO leadership and policy have placed renewed emphasis on evaluating patent claims as a whole, rather than at an overly abstract level. Current guidance instructs examiners to focus on whether claims recite a specific technological solution or improvement, as opposed to merely involving an abstract concept.

Importantly, the USPTO has clarified that many computer-implemented and AI-related processes are not mental processes simply because they involve algorithms or data analysis. Where claims describe concrete technical mechanisms and improvements, they may satisfy patent eligibility requirements even in complex software and AI contexts.

Ex parte Desjardins and Its Impact on AI and Software Claims

A key development influencing today’s eligibility analysis is the PTAB’s precedential decision in Ex parte Desjardins. In that case, the Board vacated a § 101 rejection after finding that the claims were directed to improvements in how a machine-learning model itself operates, rather than the mere application of known techniques to a new dataset.

The decision draws an important distinction for applicants:

• Claims that apply generic machine-learning techniques to a new field or business problem remain vulnerable under § 101.
• Claims that articulate specific technical improvements to model training, system architecture, or computational performance are more likely to be viewed as patent eligible.

For AI innovators, this distinction reinforces the importance of clearly describing how an invention improves underlying technology, not just outcomes or results.

Federal Circuit Decisions Continue to Shape the Boundaries

Despite evolving USPTO guidance, Federal Circuit decisions continue to reflect tension in § 101 jurisprudence. Recent opinions emphasize that improvements such as speed, efficiency, or automation alone are insufficient unless they are tied to a specific technological improvement.

This divergence underscores the importance of thoughtful drafting and prosecution strategy. Applicants must be prepared to articulate why claimed inventions improve computer functionality or technological processes, rather than merely automating or optimizing existing practices.

Drafting and Prosecution Considerations

Across software, AI, and other computer-implemented technologies, several practical considerations can strengthen a patent’s eligibility position:

• Clearly identify a specific technical problem in the specification.
• Describe how the invention solves that problem through concrete technical mechanisms.
• Explicitly articulate improvements to system performance, efficiency, accuracy, or resource usage.
• Avoid purely result-based or business-focused claim language.
• Emphasize how claim elements interact as an ordered combination to produce a technological improvement.

When addressing § 101 rejections, applicants should respond comprehensively, challenging over-generalization, demonstrating practical application, and explaining why the claims recite more than a generic computer implementation.

Looking Ahead

While patent eligibility remains a complex and evolving area of law, recent developments indicate a more favorable environment for well-drafted software and AI patents. Applicants who focus on technical improvements, detailed disclosures, and strategic prosecution practices are better positioned to navigate § 101 successfully.

Compliments of Panitch Schwarze – a member of the EACCNY