BREAKING: Section 101 Rejections Soar at USPTO; 77% of AI Tech Group's OAs Include 101 Rejections
Also: USPTO working on new 101 guidance; best practice for avoiding 101 issues at USPTO
Clause 8 Podcast
Last week - IP Fray's Florian Mueller on EU’s SEP regulation efforts:
"We should all brace for impact and should assume that something's going to come out of this, and 2025 is probably the year."
From the archives on SEPs:
Joff Wild was more optimistic about the EU proposals not becoming law BUT warned that EU Commission’s evidence-free SEP proposals risk destroying EU’s global leadership on SEP/FRAND issues
Qualcomm’s Chief IP Counsel Bob Giles on the indispensable role that SEPs play for wireless technology innovation
On next week’s episode: why a $850 billion pharmaceutical company obtains only 50 patents per year
Section 101 rejections soaring at USPTO in 2024
Voice of IP tipster lost in Alice’s not so wonderland:
While the patent bar is distracted by the flurry of problematic new proposals, the USPTO seems to be returning to its old ways when it comes to issuing lots of bad Section 101 rejections. I haven’t seen it this bad since before the 2019 PEG was issued by the USPTO. What is going on!?
Since Voice of IP has had a longstanding interest in the subject, we jumped at the chance to investigate.
Flashback: almost exactly ten years ago - on June 19, 2014 - the Supreme Court’s Alice decision upended more than 15 years of settled law with regards to the patentability of software-related innovations under §101. Unfortunately, the Federal Circuit failed to provide meaningful clarity in the years that followed. This led to widespread confusion and inconsistency among USPTO examiners, who were unrealistically expected to compare claimed inventions to those in decisions dealing with §101. This often involved examiners pigeonholing claims at issue into completely unrelated court decisions that found claims to be directed to an abstract idea. Because of the “choose your own adventure” nature of Federal Circuit’s §101 decisions, practitioners and applicants stuck with §101 rejections often found it impossible to reach any positive resolution.
Newly confirmed USPTO Director Andrei Iancu decided to finally tackle this issue - at least within the USPTO - by releasing the the 2019 Revised Patent Subject Matter Eligibility Guidance (2019 PEG). The guidance eschewed the “impractical” practice of of expecting examiners to compare claims to a growing number of court decisions involving §101. Instead, the guidance provided a workable approach for §101 analysis by USPTO examiners.
As the table above shows, the 2019 PEG led to a 10% decrease in §101 rejections from 25% in 2018 to 15% in 2020.
Is the USPTO backsliding on Section 101? Voice of IP relied on USPTO’s own data1 to find out the answer.
Interestingly, the rate of §101 rejections remained relatively steady for the next 3 years: 16% of all Office Actions had §101 rejections in 2021 and 17% of all Office Actions had §101 rejections in 2022 and 2023. However, that dramatically changed at the beginning of this year.
Section 101 rejections soar in ‘24 to pre-2019 PEG levels: in January 2024, 23% of all Office Actions had §101 rejections. As the table above shows, this high rate of §101 rejections has continued through 2024 and the average for the year is within a percentage point of overtaking the all-time high rate of 25% from the year before the 2019 PEG is issued.
77% of AI tech group's OAs include 101 rejections: Voice of IP has identified the most dramatic increase in §101 rejections in USPTO’s working group 2120, which examines patent applications related to artificial intelligence and simulation modeling. 77% of Office Actions issued by that group have a §101 rejection. This rate is actually almost 20% higher than from before the 2019 PEG guidance was issued and more than double than in 2022.
Impact: so while the USPTO is championing the importance of developing AI technologies in the US, it’s becoming significantly more difficult to obtain patent protection for AI-related inventions because of the increase in §101 rejections at the USPTO. As former USPTO Director David Kappos noted back in 2019, a constricted approach to §101 undermines investment in AI technologies.
New Section 101 guidance/proposal from the USPTO to bolster development of AI technologies?
Voice of IP has learned that the USPTO is currently working on new guidance for §101. It’s not yet clear what it will cover or what form it will take. USPTO’s recent actions in other areas will understandably lead to concerns that the USPTO is going to try to also use §101 to make it more difficult to obtain robust and reliable patent rights. However, the USPTO has an opportunity to use new guidance for §101 to show that its actions match its rhetoric regarding AI.
Currently, MPEP 2106.04(a)(1) provides a single example to examiners regarding how “a method of training a neural network” does not recite an abstract idea. Most examiners are not aware of this example. Those who become aware of it tend to treat it as a very narrow example or an aberration. This explains the high rate of §101 rejections for AI-related applications. Therefore, like the USPTO did with the 2019 PEG, the USPTO should issue new guidance that focuses on helping examiners avoid wrongly making §101 rejections for claims covering AI-related inventions.
Eli’s best practice take: the table above again highlights how ending up in the wrong art unit within the USPTO can doom an applicant’s ability to obtain the best possible patent protection in an efficient way. For example, an application landing in an art unit of WG 3690 will lead to a §101 rejection 9 out of 10 times. This will likely force an applicant to narrow their claim rights and expend resources in ways that they would not have had to do if they landed in art unit with a lower rate of §101 rejections.
Therefore, the best practice remains to draft patent applications in a way that will help those applications avoid those §101 trigger-happy art units. This will also likely lead to issued patents have a better chance of surviving.
Do you know what the USPTO is working on with regards to Section 101? Do you have any thoughts about one of those Clause 8 episodes listed above?
Please email me at eli@VoiceOfIP.com or find me on Signal at eli.92.
https://developer.uspto.gov/visualization/agency-trends-rejections-office-actions-patent-applications. The above table relies on the time period from January-June for ease of comparison with this year. However, those rates remained relatively similar over those years.