A New Era In Patent Eligibility At The USPTO: An Opportunity For Agritech?

For the last decade, securing patent protection for software inventions has been challenging. However, in recent months the United States Patent and Trademark Office (USPTO) has signaled that restrictions on patenting software-based inventions may be easing. This may present an opportunity to industries such as Agritech that increasingly rely on software and artificial intelligence (AI) and machine learning.

Agritech:

Agritech refers to the application of technology and innovation to agriculture, horticulture, and aquaculture to improve efficiency, sustainability, and yield. Tools such as AI/machine learning-enabled drones, sensors, and automated systems help producers manage resources more effectively and produce more food with fewer inputs. Particular examples of Agritech include precision farming, smart irrigation, robotic planting and harvesting, and implementation of data analytics to inform decision-making. Agritech applications can benefit agriculture by: 1) increasing productivity with less use of resources such as water, fertilizer, and pesticides; 2) reducing the impact of agricultural operations on natural ecosystems; 3) reducing chemical runoff into rivers and groundwater; and 4) improving worker safety. This can lead to greater food production efficiency, more sustainable farm operations, cost reductions, safer growing conditions and safer foods, reduced environmental impact, and improved supply chain efficiency.

Software and Agritech:

Agritech increasingly relies on software-based technologies, including: AI; the Internet of Things (IoT); robotics, sensors, and drones; satellite imagery; and precision farming techniques. AI and machine learning algorithms are capable of analyzing vast amounts of data and, properly trained, can provide helpful guidance for crop modeling, predictive analytics such as predicting crop yields and market prices, and decision support systems for optimized pest control and fertilization. Robots can be adapted to fully or partially automate tasks like planting, harvesting, and weed control, which can significantly reduce labor costs and increase productivity. Automated systems can also be adapted for waste management and livestock feeding. Sensor technology finds use in real-time monitoring of crop health, soil conditions, and weather for precision irrigation and other farm practices. Drones and satellite imagery can be used for monitoring crops and mapping fields, and for precision application of pesticides and fertilizer in place of broadcast application. The science of precision farming incorporates various technologies to optimize crop yield, including automated irrigation systems and GPS-guided tractors, seeders, harvesters, and other equipment.

Treatment of Software-Related Patent Applications:

Particularly in the fields of AI, data analytics, bioinformatics, and IoT applications, the USPTO and the courts have in recent years hampered the ability to protect software-based innovations. The claims of patents and patent applications directed to inventive and useful AI and other software-related inventions are often invalidated or rejected as per se ineligible for patenting under 35 U.S.C. §101, the statute that identifies the four patentable categories of inventions (process, machine, article of manufacture, or composition of matter). The most common reason put forth is that the claims are allegedly directed to overly “abstract” concepts rather than definable inventions, rather than for lack of novelty or as being legally “obvious” modifications of earlier inventions.

The U.S. Supreme Court Alice Decision and Its Consequences:

The high level of rejections and invalidations of software-related patent applications stems in part from the U.S. Supreme Court’s 2014 decision in Alice Corp. v. CLS Bank Int’l.¹ In that case the Court invalidated certain business method patents as having claims drawn to an abstract idea without sufficient additional features to transform the abstract idea into patentable subject matter, i.e., the mythic and elusive “something more” often pointed to by patent examiners and the courts as conveying patent eligibility to a software claim.

Leading up to the Supreme Court’s holding, the U.S. District Court for the District of Columbia invalidated Alice Corp’s claims² as statutorily ineligible for patent protection because they preempted an abstract concept, specifically the basic business or financial concept of employing an intermediary to minimize risk by facilitating a simultaneous exchange of obligations. The inclusion of a computer system that performed the functions did not salvage patentability of the claims. The case then went to the U.S. Court of Appeals for the Federal Circuit which initially reversed the district court, holding that it was not “manifestly evident” that Alice’s claims were directed to an abstract idea.³

On en banc rehearing at the Federal Circuit, in a fractured opinion⁴ a panel of 10 judges issued seven different opinions with no majority support for any one opinion. Seven of the 10 judges affirmed (for different reasons) the District Court’s ruling that the claims were ineligible. Half of the panel agreed with the District Court’s holding that the computer system claims were not patent-eligible, and the other half did not.

The Federal Circuit en banc panel did not enunciate any single standard for determining eligibility of computer system claims. The plurality opinion set out a test for patent eligibility which required first identifying the abstract idea or fundamental concept applied by a claim, and then determining whether the claim preempted the abstract idea or fundamental concept. A four-step analysis proposed for this determination required: 1) determining whether the claimed invention fits within one of the four statutory classes of patentable inventions; 2) determining whether the claimed subject matter risks “preempting an abstract idea;” 3) identifying the idea supposedly at risk for preemption by defining “whatever fundamental concept appears wrapped up in the claim;” and 4) analyzing the “inventive concept,” that is, determining whether the claims reflect genuine, human contribution to the claimed subject matter that contains “additional substantive limitations that narrow, confine, or otherwise tie down the claim so that, in practical terms, it does not cover the full abstract idea itself.” In principle, the analysis had the laudable goal of avoiding issuing claims that preempt entire abstract concepts rather than being directed to specific inventions.

The case was then appealed to the U.S. Supreme Court. Relying on its earlier decision in Mayo v. Prometheus,⁵ the Court reiterated the two-step analysis set out in Mayo which required: 1) determining whether a claim involves an abstract idea such as an algorithm, method of computation, or other general principle; and 2) determining whether the claim provides “something extra” that defines an “inventive concept.” The Court held that without that inventive concept, a claim involving an abstract idea cannot be patent eligible. The Court also held that a mere instruction to implement an abstract idea on a computer cannot impart patent eligibility, and that mere recitation of a generic computer likewise cannot transform a patent-ineligible abstract idea into a patent-eligible invention. Unfortunately, while the Court addressed the most basic issue in the case (statutory ineligibility of business method patents), it provided little to no guidance on how that analysis should be conducted and/or what test(s) should be applied.

The Alice patents were directed towards business methods and did not mention computer software. However, the Alice/Mayo analysis has since been applied to software-based inventions in general. The Supreme Court’s limited guidance and reluctance to identify specific applicable tests for determining what is and is not statutorily patent-eligible led to extraordinary inconsistency in the analysis of statutory eligibility of software claims. This resulted in years of uncertainty on how to draft software claims to meet the bar for patent eligibility, and in the opinion of many unduly restrictive analysis of statutory patent eligibility of software claims by patent examiners and the courts. As an example (although the rate has since moderated), in the year immediately following the Alice decision the Federal Circuit Court of Appeals invalidated more than 90% of software patents brought before them for review under Section 101 as being directed to ineligible subject matter.

A Shift in the USPTO’s Stance on Eligibility Under 35 U.S.C. § 101:

There is some good news for software patent applicants and patentees. In recent months the USPTO has signaled a shift in how the Office will address statutory eligibility in examining software and other fundamental technologies going forward, particularly for AI-based inventions. In August 2025, the Office issued a Memorandum⁶ clarifying how patent examiners are to apply the existing subject matter eligibility guidance set forth in the Manual of Patent Examining Procedure (MPEP). The Memorandum was directed to Technology Centers 2100, 2600, and 3600 (Computer Architecture Software and Information Security, Communications, and eCommerce), which are the Technology Centers that examine the majority of software applications.

The Memorandum does not expressly countermand any prior USPTO guidance on examination of software-based inventions, and is instead styled as a reminder to patent examiners of how to apply such guidance. However, the Memorandum presents significantly more applicant-friendly guidelines to patent examiners and in doing so provides useful guidance to patent applicants on avoiding or rebutting Section 101 rejections. In particular, the Memorandum places limits on examiner invocation of the mental process-abstract idea exception to patent eligibility, emphasizing that if a claim limitation cannot practically be performed by the human mind then the limitation should not be classified as a mental process. Specific to AI, the Memorandum states that “Claim limitations that encompass AI in a way that cannot be practically performed in the human mind do not fall within this grouping [mental process].” Examiners are reminded that claims that merely “involve” an abstract idea as opposed to “reciting” and therefore potentially preempting the abstract idea do not fall under the abstract idea exception. Examiners are also reminded that claims must be analyzed “as a whole” for statutory eligibility, rather than considering individual limitations “in a vacuum, completely separate from the recited judicial exception” while substantially ignoring other limitations as being “recited at a high level of generality” as was often previously examiner practice.

The Memorandum further states that AI claims may be eligible if they reflect an improvement to computer technology or another area of technology, and reminds examiners to consult the specification to determine whether such improvements are disclosed. Examiners are cautioned against oversimplification of claimed subject matter to support a rejection of a claim as directed to a mental process-abstract idea. Finally, the Memorandum states that examiners should make a rejection of claims on statutory ineligibility grounds only when “it is more likely than not (i.e. more than 50%) that the claim is ineligible.” Examiners are cautioned not to issue rejections simply because they are uncertain that a claim is eligible. This “preponderance” standard should prove favorable to applicants if applied by patent examiners.

Newly appointed USPTO Director John Squires has issued further indications that analysis of eligibility under section 101 has entered a more applicant-friendly era. In Ex parte Desjardins et al.⁷ the USPTO Appeals Review Panel (ARP) including Director Squires, Acting Commissioner Wallace, and Vice Chief Judge Kim sua sponte vacated a Patent Trial and Appeal Board (PTAB) rejection of AI-related claims under Section 101. The Desjardins decision stated that the eligibility analysis required the PTAB to “evaluate the significance of the additional elements relative to the invention,” and noted that “the ultimate question is whether the [judicial] exception is integrated into a practical application” (citing the Manual of Patent Examining Procedure §2106.04(d)(II); emphasis added). The ARP further stated that “Categorically excluding AI innovations from patent protection in the United States jeopardizes America’s leadership in this critical emerging technology. Yet, under the [PTAB] panel’s reasoning, many AI innovations are potentially unpatentable – even if they are adequately described and nonobvious – because the panel essentially equated any machine learning with an unpatentable ‘algorithm’ and the remaining additional elements as ‘generic computer components,’ without adequate explanation … Examiners and panels should not evaluate claims at such a high level of generality.”

Finally, the ARP stated that “[a]t the same time, the claims at issue stand rejected under [35 U.S.C.] §103 [obviousness]. This case demonstrates that §§102 [anticipation], 103, and 112 [indefiniteness, written description, and enablement] are the traditional and appropriate tools to limit patent protection to its proper scope. These statutory provisions should be the focus of examination” (emphasis added). Reflecting prior opinions of many in the industry, the patent bar, and certain courts, the ARP appears to suggest that patent-eligibility should remain a low bar and a threshold test, rather than as previously a significant bar to patentability.

Takeaways:

Of course, guidance issued and actions taken by the USPTO are not binding on any Article III court. However, patent applicants in the Agritech sphere can expect that patent examiner subject matter rejections of properly crafted claims directed to software, AI, and machine learning-based inventions should decrease in the short-term. To maximize their likelihood of avoiding rejections under Section 101, patent applicants should take care to emphasize that the inventions as claimed apply any “abstract idea” involved, rather than merely reciting it. Further, applicants should ensure that patent application specifications explain in detail any and all technical advantages of as many different features of the disclosed invention as possible to make the case that the claims recite an improvement in the subject Agritech technology involved.

Footnotes

1. Alice Corp. v. CLS Bank Int’l, 573 U.S. 208, 134 S. Ct. 2347 (2014).

2. CLS Bank Int’l v. Alice Corp. Pty, Ltd, 768 F. Supp.2d 221 (D.D.C. 2011).

3. CLS Bank Int’l v. Alice Corp. Pty, Ltd, 685 F.3d 1341 (Fed. Cir. 2012).

4. CLS Bank Int’l v. Alice Corp. Pty, Ltd, 717 F.3d 1269 (Fed. Cir. 2013).

5. Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. 66 (2012).

6. Reminders on evaluating subject matter eligibility of claims under 35 U.S.C. 101, Charles Kim, Deputy Commissioner for Patents, August 4, 2025.

7. Appeal 2024-000567 (September 26, 2025).

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.