Ex Parte Desjardins: The Future of AI Patents Just Got Brighter

Estimated reading time: 8 minutes

Summary

For years, the phrase “abstract idea” haunted innovators in artificial intelligence and software. It became a pervasive barrier, blocking truly inventive technologies from obtaining patent protection. Countless brilliant minds watched the U.S. patent system dismiss their creations as ineligible, stifling innovation and creating immense uncertainty. But now, a landmark decision from the very top of the U.S. Patent and Trademark Office (USPTO) offers a powerful beacon of hope. Ex parte Desjardins signals a definitive shift, redefining the landscape for the future of AI patents and paving the way for a new era of innovation.

Prefer to watch instead? Check out the video below.

How the US Patent Office Just Changed Patent Examination For Software and Computer Related Innovation

I’m Mark Trenner, a patent attorney dedicated to helping AI startups and tech founders protect their most critical inventions. When a decision like Desjardins emerges, I meticulously analyze every word. It directly changes the strategies we employ to secure strong patents for our clients. This decision provides a robust new argument, empowering inventors to protect their groundbreaking work more effectively than ever before.

The Persistent Problem: Section 101 as an Unfair Gatekeeper

For those outside the patent world, Section 101 of the Patent Law seems deceptively simple. It states that an invention must fall into one of four categories to be patent-eligible: a process, machine, manufacture, or composition of matter. Historically, patent examiners interpreted Section 101 to exclude “abstract ideas” – fundamental concepts like mathematical formulas or basic economic principles – from patent protection. The intent was to prevent monopolies on foundational knowledge.

However, its application to software and AI became notoriously inconsistent and overly broad. Examiners frequently applied the “abstract idea” label to inventions that demonstrably solved real-world technical problems and offered tangible improvements to computer functionality. This created a frustrating, almost impossible scenario for innovators in the computer space. They developed innovative systems, poured resources into R&D, only to face rejections based on an often-vague and subjective interpretation of patent eligibility. The system was failing to adapt to the very technologies it should protect.

Ex parte Desjardins: A Specific Solution, An Abstract Rejection

Let’s dive into the specifics of Ex parte Desjardins. This wasn’t a case about some generalized AI concept. The patent application described a concrete, highly technical method for training machine learning models. The technology promised real-world benefits and offered demonstrable technical advantages:

• It helped models “use less of their storage capacity,” a crucial improvement for efficiency and scalability.
• It enabled “reduced system complexity,” simplifying the architecture and maintenance of AI systems.
• It allowed models to “learn new tasks in succession whilst protecting knowledge about previous tasks,” a direct address to the persistent challenge of catastrophic forgetting in continual learning.

These are not abstract goals. They represent tangible improvements to the core functionality and efficiency of AI systems. Yet, the patent examiner initially rejected the claims under Section 101, labeling them an “abstract idea.”

The applicant, committed to protecting their innovation, appealed to the Patent Trial and Appeal Board (PTAB). Unfortunately, the PTAB not only affirmed the examiner’s initial rejection but introduced a new reason for the Section 101 rejection. This scenario epitomized the ongoing struggle for AI inventors: even after demonstrating clear technical merit, the basic eligibility of their invention remained a moving target.

The Turning Point: The Appeals Review Panel Steps In

This is where the story of Ex parte Desjardins transforms from a frustrating account of patent challenges into a landmark decision for the future of AI patents. The case escalated to a special Appeals Review Panel (ARP). This was no ordinary panel; it included the new USPTO Director John Squires, Acting Commissioner Valencia Martin Wallace, and Vice Chief Judge Michael Kim. The Director’s personal involvement signaled the agency’s intent to make a definitive statement on this critical issue.

And a definitive statement they made. The ARP overturned the PTAB’s Section 101 rejection in this case.

Their reasoning was pivotal. The panel drew heavily on the Federal Circuit’s 2016 decision in Enfish, LLC v. Microsoft Corp. The Enfish case established a crucial principle that directly benefits software and AI inventions: advancements in computer technology “by their very nature, may not be defined by particular physical features but rather by logical structures and processes.” Essentially, an invention does not require new hardware to qualify as patent-eligible. If it provides a specific, technical improvement to computer functionality through software or logical structures, it can overcome the abstract idea hurdle.

In Desjardins, the ARP found the applicant’s claims fit squarely within this Enfish framework. The method for training machine learning models was not a mere abstract concept. It constituted a specific, technical improvement that enhanced the capabilities of AI systems, making them more efficient and effective. This was a direct improvement to the functionality of the computer itself, not just an application of an abstract idea.

Beyond Section 101: Focusing on the “Right” Questions

Perhaps one of the most significant aspects of the Desjardins decision lies in its broader implications for patent examination and the future of AI and computer software patents. The ARP explicitly stated that while the claims were eligible under Section 101, they still stood rejected under Section 103 for obviousness.

The ARP used this fact to drive home a critical point: Sections 102 (novelty), 103 (non-obviousness), and 112 (written description) are the “traditional and appropriate tools to limit patent protection to its proper scope.”

Consider this fundamental distinction. Section 101 should serve as a basic threshold test for what kinds of things can be patented. Once an invention passes that initial screen, the real work of examination (determining if it is new enough, non-obvious enough, and described well enough) should fall to Sections 102, 103, and 112. For too long, Section 101 functioned as an overly aggressive, often arbitrary, gatekeeper, preventing even truly innovative AI and software patents from ever reaching the substantive examination stage. This created a situation where the merit of the invention itself was rarely fully evaluated.

The Desjardins decision requires examiners to shift their focus. Instead of endlessly debating the “abstractness” of a technology that clearly provides technical improvements, they should instead examine whether the invention is truly novel and non-obvious, and whether the claims adequately describe it. This represents a more rational, effective, and ultimately fairer approach to patent examination for complex, cutting-edge technologies like AI and computer-related inventions. It allows the system to focus on the technical merits, which is where the value of innovation truly lies.

What This Means for AI Innovators and Patent Attorneys: A Brighter Future

For founders, inventors, and fellow patent attorneys working with AI patents, Ex parte Desjardins ushers in a renewed sense of optimism and clarity. It profoundly impacts the future of AI patents.

1. Stronger Arguments Against Section 101 Rejections: This decision provides powerful new ammunition. If your AI invention offers concrete technical improvements to computer functionality, you now wield strong precedent from the highest levels of the USPTO to counter Section 101 abstract idea rejections. This equips you with a formidable legal basis to advocate for your invention’s eligibility.
2. Focus on Technical Specificity: The case reinforces the critical importance of drafting patent applications that clearly articulate the specific technical problem your AI solves and the precise technical solution it provides. Vague or overly generalized claims remain vulnerable. Your application must tell a compelling technical story.
3. Shifting Examination Landscape. Desjardins will not eliminate all Section 101 challenges overnight. But, it unequivocally encourages examiners to apply the eligibility framework more appropriately and consistently. This means a greater likelihood that the USPTO will evaluate your AI innovation on its actual inventive merits. Examination should focus on novelty and non-obviousness, rather than prematurely dismissing it on eligibility grounds.
4. A Win for Innovation: Ultimately, this decision is a resounding victory for the entire AI ecosystem. This case provides more clarity and a more reasonable approach to patent eligibility. It reduces uncertainty and actively encourages further investment and development in AI technologies. When inventors can more reliably protect their intellectual property, they are more likely to take risks, push boundaries, and bring transformative technologies to market. This fuels economic growth and technological advancement.

Navigating the New Landscape of AI Patents

The patent landscape for AI constantly evolves, but decisions like Ex parte Desjardins represent monumental progress. If you develop breakthrough AI technology, work with experienced patent counsel who understand these nuances. Find an attorney who can craft strategies tailored to the latest legal developments becomes more critical than ever. You need to ensure your applications are not just technically sound. Your application needs to be legally robust against potential challenges under Section 101 of the Patent Law.

Do not let historical Section 101 hurdles deter your innovation. The game is changing, and with the right strategic approach, protecting your groundbreaking AI patents is more achievable than ever before. The future of AI patents looks considerably brighter, and proactive inventors stand to gain the most.

You may also want to read:

• AI Patents Explained for Startups and Tech Founders
• What is an Office Action from the Patent Office?
• What are the parts of a Patent Application?
• What is Patent Information and Why is it Important?
• Patent Examiner Rejections