“The entirety of the patent bar can agree that the Alice decision, along with the U.S. Court of Appeals for the Federal Circuit’s 2008 en banc decision in In re Bilski…have left things clear as mud.”
As an update to my previous posts from 2017, 2019, 2020, March 2021, August 2021, and 2022, it has now been almost nine years since the U.S. Supreme Court’s 2014 Alice Corp. v. CLS Bank decision. Yet the debate still rages over when a software (or computer-implemented) claim is patentable versus being simply an abstract idea “free to all men and reserved exclusively to none” (as eloquently phrased over 74 years ago by then-Supreme Court Justice Douglas in Funk Bros. Seed Co. v. Kalo Inoculant Co.).
Further, it has been 12 years since famed venture capitalist Marc Andreessen wrote the influential and often-quoted op-ed piece in the Wall Street Journal titled “Why Software Is Eating the World.” Today, the digital transformation where software is “eating the world” is undeniable. Artificial Intelligence (AI), the Metaverse, Web3, cloud, gene editing, autonomous driving, quantum computing, and “green tech” dominate the technology news headlines and technology trend forecasts – all heavily reliant on software-related innovation – [Forbes] [Gartner] [World Economic Forum], but we are still without concrete guidelines for software-related patenting.
Let’s look at some facts and figures:
- According to the NY Federal Reserve, computer engineering and computer science represent 2 of the top 5 highest paying majors in the USA [CNBC].
- Software engineer is the number one in-demand job of 2023, where the median salary is US$120,730, and the U.S. Bureau of Labor Statistics expects 26% growth in this profession over the next 10 years [S. News & World Report].
- US SaaS VC investment hit US$90 billion in 2021, the highest on record, with over 263 US SaaS VC deals greater than US$100 million – 3x the total last year and 7x the total in 2015 [SVB].
- In 2022, global M&A activity in the Software and Internet Services space reached US$973 billion, representing over 10,544 deals [IMAA] – this includes the acquisitions of VMware, Citrix, Anaplan, and Zendesk [CIO].
- In the 2020-21 academic year, 34,164 Associate’s degrees, 90,775 Bachelor’s degrees, 54,174 Master’s degrees, and 2,572 PhDs were awarded in the U.S. in computer and information sciences [National Center of Educational Statistics].
- Gartner estimates that total 2023 global IT spending will reach US$4.5 trillion in 2023, an overall increase of 2.4% from 2022, but with the software and IT services segments projected to grow 9.3% and 5.5% in 2023, respectively, while the devices segment is forecast to decline 5.1% [Gartner].
So, given the above, we again sought to determine what percentage of issued patents by U.S. Patent and Trademark Office (USPTO), as well as the European Patent Office (EPO) and the China National IP Administration (CNIPA) are “software-related”?
To define what patents are “software-related,” we utilized the same methodology employed by the United States Government Accountability Office in a 2013 report to Congress. That report relied on certain United States Patent Classification (USPC) classes and subclasses of applications most likely to include software-related claims as selected by expert advisors to the USPTO. Now, with the assistance of IP services firm Clairvolex (and after converting from the USPC system to the Cooperative Patent Classification (CPC) system), data was pulled from the PatSeer Global Patent Database. The results? In 2022:
- 63.5% of issued U.S. utility patents were “software-related” (relatively unchanged from the 63.1% in 2021)
- 49.6% of granted EPO patents were “software-related” (relatively unchanged from the 49.4% in 2021)
- 43.5% of granted Chinese patents were “software-related” (up 8% from 40.2% in 2021)
- The top 15 software-related granted U.S. utility patent assignees were:
Despite the significance of the above numbers, the entirety of the patent bar can agree that the Alice decision, along with the U.S. Court of Appeals for the Federal Circuit’s 2008 en banc decision in In re Bilski (concluding that the “machine or transformation” test is the governing test for determining patent eligibility of a process under Section 101 of the Patent Act), have left things clear as mud. In direct response to this uncertainty, the USPTO issued its 22-page (with 87 footnotes and 42 mentions of “computer”) 2019 Subject Matter Eligibility Examination Guidance based on the Alice decision and other Supreme Court and Federal Circuit opinions on patent eligibility. That guidance document has now been included in Section 2106 of the Office’s latest edition of the Manual of Patent Examining Procedure (MPEP e9 r7.2022) and can quickly be summed up as:
Further, and without getting too much into “Stats 101” and debates over lies, damn lies, and statistics, I leave you with these observations (keeping in mind that over half of all issued U.S. patents over the past decade have been software related):
- “High-tech” (i.e., software, hardware, and networking) patents account for nearly 70% of US patent litigation [Unified Patents];
- Nearly two-thirds of patents challenged at the USPTO’s Patent Trials and Appeals Board (PTAB) in an IPR have a preceding district court case [RPX];
- No USPTO Technology Center has a survival rate at the PTAB in IPR, CBM, and PGR cases above 44% [Finnegan Henderson]; and
- The Federal Circuit (partially or completely) has historically affirmed the PTAB in over 83% of IPR, CBM and PGR appeals [Finnegan Henderson].
In light of all this, maybe some legislative help is on the way? Well, last summer U.S. Senator Thom Tillis (R-NC) released draft legislation titled the Patent Eligibility Restoration Act of 2022 [S.4734] to address the Section 101 problem. In the accompanying press release, he summarily and correctly stated:
“Unfortunately, our current Supreme Court’s patent eligibility jurisprudence is undermining American innovation and allowing foreign adversaries like China to overtake us in key technology innovations. … Unfortunately, due to a series of Supreme Court decisions, patent eligibility law in the United States has become confused, constricted, and unclear in recent years. … As of 2021, all 12 judges of the United States Court of Appeals for the Federal Circuit have lamented the state of the law. … [R]eforms are necessary to restore the United States to a position of global strength and leadership in key areas of technology and innovation, such as medical diagnostics, biotechnology, personalized medicine, artificial intelligence, 5G, and blockchain.”
Needless to say, Senator Tillis’ proposed legislation went nowhere in the 117th Congress. So, what will this new 118th Congress do to reform Section 101 of the Patent Act as it seems to be preoccupied with fighting the asleep culture wars? We will see. Until then, when it comes to patenting software, the struggle is still real.
Join the Discussion
14 comments so far. Add my comment.
AnonMarch 31, 2023 01:33 pm
Yes, Lab, I am serious.
It is not that I have closed my ears to the litany of small and disruptive firms bemoaning the rather evident anti-patent environment.
Such is not something that I have ever denied.
Rather, I merely point out that (despite that environment, that) the facts presented in this article STILL show a massive upswell in those seeking to protect the single most abundant form of innovation today.
As I have posted here and elsewhere, we are in the fifth Kondratiev wave, so it is simply not a surprise the type of innovation that is so often sought to be protected — yes, even in the face of the onslaught from both the courts and Efficient Infringers, and — to close with your own admission — “in the light of the importance of computer technology, as emphasized in the above article.”
My post merely emphasizes what is emphasized in the above article, and YOUR assertion stands in contrast to the facts presented (again, in spite of the difficulties).
Lab JedorMarch 30, 2023 07:52 pm
Really Anon? Have you missed the litany of small, big and medium sized companies who have lost their supposedly secure IP property rights on computer implemented inventions?
The last big win for software patents was in NPT v. Research in Motion. Big tech decided after that decision, “never again”. And with stunning ease and great success they succeeded. AIA, eBay, Alice and others all contributed in undermining software related patents.
Perhaps you have (for me very welcome) facts why a company and especially a start-up should file a patent application for market critical technology, and having a better than 75% chance to avoid invalidation and the ability to prevent an infringer from using the claimed invention. There is no patent attorney who will even try to guarantee that at a 50% level. Well, I have to qualify that. No patent attorney I know of. That means you have a better chance of winning at roulette.
A successful patent, like the RSA or the NPT patent, will not be successfully asserted any time soon, if it even survives an IPR or if it is even issued. That is bizarre of course in the light of the importance of computer technology, as emphasized in the above article.
OstrichesMarch 30, 2023 05:35 pm
Patent Eligibility jurisprudence is absurd, illogical Doublespeak at every single level – from the Examining Corps to the U.S. Supreme Court. As is the “recent” patent legislation under the AIA.
When I read Office Actions and caselaw for purposes of forming a reasoned and cogent response regarding eligibility, I often lament what I am reading and am reminded of the saying, “Answer not a fool according to his folly, lest thou also be like unto him.”
I think, maybe, after 20+ years of prosecution it is time for me to stop banging my head against the wall and find another line of work.
Logical arguments and reasoning cannot be appreciated by illogical and unreasonable beings – that’s assuming they are even reading what is written.
This is what it has become.
AnonMarch 30, 2023 03:14 pm
Aside from any other aspects of your dialogue,
“Not only that, in computer implemented inventions the patent system probably has very limited benefits for any company and may play against you.”
is simply not borne out by the facts at hand — which is the emphasis of this article.
If one is going to make that assertion, then one probably wants to be able to provide some reasoning for the factual data presented.
Lab JedorMarch 30, 2023 01:19 pm
Curious. The answer is in you own observations. Once you are big enough to use the system to your benefit, by having sufficient funding, a positive cash flow and a significant market share, the patent system starts to work to your advantage.
You have enough market power, so you don’t have to worry too much about being knocked out of business by other patent owners. Furthermore, you can tie-up start-ups that rely on patents in knots by endless IPRs and allegations of abstract ideas, while cleverly including their inventions. But you can be destroyed by new technology you have no access to.
It suggests that you preferably start using patents when the system can be used to your advantage. When you are big. But those benefits of the patent system apply less and less to start-ups.
Not only that, in computer implemented inventions the patent system probably has very limited benefits for any company and may play against you.
You keep on repeating the enormous number of patents the big boys have. But IBM has already decided that the numbers do not play a role anymore. And I bet you that Google with all its thousands of patents has no IP to knock out ChatGPT. In fact ChatGPT would be crazy to teach Google what the critical advantage is they have.
It is fascinating to see how AI is developing market wise. Google is trying to catch up to ChatGPT with Bard, and with limited results. The whole chatbot space is developing so rapidly and having such an enormous potential, that is seems unwise to give away any advantage you have by publishing your critical technology. Furthermore, any patent application in this space has the risk that it may receive a secrecy order. Stopping commercial development right in its tracks.
I am not in favor of this. I believe it is better to have a patent system that provides good protection and forces competitors to do work-arounds. But the current system does not work like that. It is geared to invalidate patents, and thus allowing, no promoting, competitors to infringe.
Any company that has critical software that can be kept secret and cannot easily be copied would be nuts to publish their crown jewels, to use mixed metaphors. It probably takes no more than 5 years to knock out the competition with proprietary AI software. We see the process unfolding in real-time.
CuriousMarch 30, 2023 11:17 am
Already, unicorns (private businesses whose valuation exceeds $1B), who can well-afford to patent, are choosing not to, especially those in areas like ecommerce, Internet, gaming and software where patents have diminishing impact.
And when the big boys (with 10s of thousands of patents) come knocking on the door of these unicorns asking them to take a license (because they have patents in areas like ecommerce, Internet, gaming and software along with AI and anything else you can think of), these unicorns will be regretting their short-sighted behavior.
Not hard to invalidate a patent or a handful of patents. Entirely cost-prohibitive to invalidate 50 or 500 or 5000 patents.
And in the sport of kings, you have to pay to play. And if you don’t pay one way, you’ll definitely be paying another way.
Pierce MooneyMarch 29, 2023 11:55 pm
More great software patent information from Mr. Millien as usual! Thank you sir.
Bruce BermanMarch 29, 2023 01:09 pm
Thank you, Ray, for another incisive article about software patents. Unfortunately, despite the significant number of software related patents, they seem to have less meaning than ever.
As you aptly point out: “Today, the digital transformation where software is “eating the world” is undeniable. Artificial Intelligence (AI), the Metaverse, Web3, cloud, gene editing, autonomous driving, quantum computing, and ‘green tech’ dominate the technology news headlines and technology trend forecasts – all heavily reliant on software-related innovation.”
It’s difficult to believe that the Congress and the courts are refusing to step up and provide a clearer path to ownership. Already, unicorns (private businesses whose valuation exceeds $1B), who can well-afford to patent, are choosing not to, especially those in areas like ecommerce, Internet, gaming and software where patents have diminishing impact.
Trade secrets would appear to be the next logical step, even though the FTC wants to eliminate non-disclosure agreements. The free-fall in IP protections is starting to look like free-for-all, and that is dangerous on many levels, even if the public and legislators refuse to see it.
CuriousMarch 29, 2023 12:53 pm
We are going to ask SCOTUS to review our situation since we came this far.
I wish you the best of luck. However, while SCOTUS may be the one place you could get real satisfaction (the CAFC was never going to give you it), if I were you, I would focus your energies at the USPTO. Your chances of having cert granted is practically zero.
AddyMarch 29, 2023 10:12 am
Lab makes a great point that many companies overlook as part of an overall IP strategy – sometimes its better to keep an innovation a secret rather than publish it in a patent. This is why business and technical leaders need to be involved in these decisions, rather than leaving them to the Legal department.
Curious highlights the need to find competent patent counsel to develop software innovation based patent applications, and this is an important point that seems to be overlooked. It’s not the fault of the “patent system” that software patents are killed in IPR or invalidated over 101. There are successful enforcement of such patents. It starts with how the patent was written and prosecuted – this provides all the reasons you need to understand why a software patent died in IPR or was invalidated due to 101. The “patent system” was doing its job.
concernedMarch 29, 2023 06:46 am
We are going to ask SCOTUS to review our situation since we came this far.
However, if I ever think a patent is valuable, please call my wife and ask her to commit me to an institution.
CuriousMarch 28, 2023 07:16 pm
This article made me really wonder if it is even useful anymore to try to patent computer implemented inventions.
It is. However, you need to understand the golden rule — he who has the gold makes the rules. In this instance, the big companies have the gold and the rules favor them.
If you are a small company and Samsung, for example, infringes one of your few patents. You could try to sue them. The likely end result is that Samsung will IPR the patent or invalidate it under 101. Samsung will likely countersue you as well for infringement. However, instead of coming after you with a single patent, they’ll assert 200 patents. Could luck invalidate 200 patents via IPRs or in litigation — NOT GOING TO HAPPEN.
Once a company’s patent portfolio gets big enough, the patents become unchallengeable. Sure, a competitor could conceivably challenge a handful of this company’s patent portfolio, but if the company asserts 10x or 100x the handful, it really doesn’t matter.
Inventions in areas such as AI, cryptography, protein design methods and other highly complex areas in many regards are hard to reverse engineer when kept inside an organization.
However, inventions of this ilk are usually very high-tech and the researchers involved like to publish (don’t ask me how I know). As such, keeping the inventions secret is not necessarily a viable solution.
Pro SayMarch 28, 2023 06:51 pm
Bottom line: Protect not. Disclose not.
Congress. SCOTUS. CAFC. PTAB.
Look at what you’ve wrought.
Innovation created on the shoulders of those who’ve come before . . . is impossible . . . without shoulders.
Lab JedorMarch 28, 2023 06:33 pm
This article made me really wonder if it is even useful anymore to try to patent computer implemented inventions.
Inventions in areas such as AI, cryptography, protein design methods and other highly complex areas in many regards are hard to reverse engineer when kept inside an organization. It seems more attractive to offer products and services derived from these inventions, but keep the inventions themselves secret.
The inability to protect a software related invention by a patent and the high invalidation rate at the PTAB, suggest that filing a patent application only serves to educate competitors without providing much protection.
In cryptography, publication of details of computer routines only educates attackers. Offering cryptography services with secret server based software limits attackers to leverage published weaknesses.
For a promising technology to finally do true rational design of proteins, it appears to be more beneficial to keeping the software secret and perhaps only patent the resulting molecules as individual products.
Highly complex AI Chatbots are difficult to reverse engineer from only (highly controlled) interaction. This while their existence appear to provide a tremendous commercial advantage. It seems wise to keep the critical technology of those Chat applications a secret and not make the mistake of educating your competitors.
Those considerations nowadays probably start to play a greater role in leading technology companies and start-ups in decisions to apply for patent protection. I don’t think numbers are available. But lack of growth in patent application while invention effort continue to grow appears to suggest that inventors are reconsidering their filing strategies.
As a VC, instead of requiring IP protection by patents of a start-up, it would probably be wiser to require to keep critical technology a secret.
It is not because I am anti-patent. On the contrary. But ultimately one has to draw the consequences of existing conditions.