The patent eligibility of artificial intelligence (AI) and other software-related inventions has been a hotly contested subject for decades. The federal statute 35 U.S.C. § 101 establishes the categories of subject matter that are eligible for patent protection – any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, is patent eligible. However, judicially established exceptions to these categories – abstract ideas, laws of nature, and natural phenomena – are held to be patent ineligible unless they are sufficiently well integrated into a practical application. There is a long line of U.S. case law that considers what constitutes a patent-ineligible “abstract idea” versus a patent-eligible practical application of an abstract idea. Ultimately, some software-related inventions are considered patent eligible while others are considered ineligible. The dividing line is not always clear. Newly sworn-in Director of the U.S. Patent and Trademark Office (USPTO), John Squires, chose to address the topic of patent eligibility under 35 U.S.C. § 101 for AI-related inventions in one of his first actions as Director. Director Squires authored a review panel decision vacating a decision by the Patent Trial and Appeal Board (PTAB) in a patent application for a machine learning-related invention1. The PTAB decision rejected the claimed invention as directed to a patent ineligible abstract idea. Director Squires criticized the original PTAB analysis of the claimed invention as “essentially equating any machine learning with an unpatentable ‘algorithm’ and the remaining additional elements as ‘generic computer components’” used merely as a tool. Director Squires instead found that the claims at issue include limitations that reflect a patent eligible technical improvement in how machine learning models are trained using multi-task learning. He found the claims at issue to be directed to more than just a patent ineligible abstract algorithm divorced from any practical application. Director Squires’ decision arguably went further than just addressing the claims at issue to specifically signal that he is taking a pro-patent stance for AI-related inventions. He warned that “[c]ategorically excluding AI innovations from patent protection in the United States jeopardizes America’s leadership in this critical emerging technology.” Moreover, he directed USPTO personnel that “§§ 102, 103 and 112 [the statutory novelty, non-obviousness, written description, and enablement requirements] are the traditional and appropriate tools to limit patent protection to its proper scope. These statutory provisions [and not § 101] should be the focus of examination.” The decision followed Director Squires’ first official act earlier that week when he held a ceremonial signing for two patents. He signed one patent directed to blockchain technology and another directed to antibody-based diagnostics. Director Squires appears to have deliberately selected two technology areas that have often faced challenges under § 101 in front of the USPTO in recent years. In remarks made at the signing that suggest Director Squires’ deliberateness, he stated that “the U.S. Patent Office is open for business, especially for the technologies of tomorrow.” Based on where Director Squires has focused his early actions as Director and his stated pro-patent perspective, it appears that patent applicants may face fewer obstacles to patenting under § 101 during Director Squires’ tenure than they have in the recent past, especially for AI-related inventions. It remains to be seen how this approach will play out at the U.S. Patent Office in the coming years and whether the case law will follow. However, these signals from the new Director, coupled with recently increased expectations from regulators for transparency and disclosure for certain AI products, may cause companies to reevaluate their IP strategies for protection of their AI products. For example, in the past, a company may have relied more heavily on trade secret protection of its AI- or other software-related products, especially where it was considered unlikely that any technical details about the software would need to be made public and where it was less certain that a patent would satisfy the subject matter eligibility requirement of § 101. However, there may now be more of a shift towards seeking patent protection of certain AI products and away from attempting to maintain trade secrets for those products if there is higher confidence that patents on those products will be found to satisfy the subject matter requirement and/or if it is known that technical details of the products will need to be made public to satisfy more stringent regulatory disclosure requirements. Ex parte Desjardins, No. 2024-000567 (P.T.A.B. September 26, 2025). ↩︎