Artificial intelligence (AI) has developed swiftly over the past decade. As consumers and industries have requested new ways to improve and simplify daily tasks, technology companies have provided phone applications, software solutions, and devices. Accordingly, legal commentators have noted that there are an increasing number of AI-related patent applications being filed in the United States Patent and Trademark Office (PTO).1 Moreover, courts have narrowed and refined the Alice/Mayo two-step process for determining software-related subject matter eligibility.2 Based in part on such jurisprudence, the PTO recently announced its revised guidance on examining subject matter eligibility under 35 U.S.C. § 101 as well as new guidance on examination procedures under 35 U.S.C. § 112.3 With the increase in AI-related patents and the ambiguity regarding the definition of AI-related technologies, the issue of AI patentability has become increasingly scrutinized in the legal community. Reflecting on the patent law developments in 2018 and early 2019, we are left to wonder: what is the outlook for patentability for AI-related technologies going forward?
How does the PTO currently define AI-related technologies?
The term “AI” appears to encompass a number of technology classifications relating to computers and data systems. Indeed, the PTO defines it in one of its technology classes – Class 706: Data Process – Artificial Intelligence.4 The PTO notes that Class 706 is a “generic class for artificial intelligence type computers and digital data processing systems and corresponding data processing methods and products for emulation of intelligence (i.e., knowledge based systems, reasoning systems, and knowledge acquisition systems); and includ[es] systems for reasoning with uncertainty (e.g., fuzzy logic systems), adaptive systems, machine learning systems, and artificial neural networks.”5 Indeed, since AI as the PTO defines it, appears to include broad categories of technologies, it is not surprising that such technologies may not necessarily be contained all in one art unit.
The PTO has defined Class 706 to include AI-related technologies, but as noted above, AI includes broad categories, which can encompass multiple art units within a class. The art units also have a range of overlap. Legal commentators have already noted how AI-related technologies have been shown to encompass a number of art units.6 Although in an ideal world, each art unit would have consistent outcomes, since different art units are assigned to different examiners at the PTO, different art units may have different rates of allowance. Thus, with respect to AI patent applications, the art unit assigned to the application may affect whether the application is ultimately allowed or rejected. The patentability of an AI-related application may depend on the careful wording of the specification and claims, which will determine the application’s art unit.
Moreover, the PTO may be paying more attention to AI-related technologies pursuant to the Federal Circuit’s interpretation of Electric Power Group, LLC v. Alstom S.A. in 2016.7 In Electric Power Group, claims directed to monitoring and reporting on the performance of an electric power grid were held ineligible under § 101 as a result of being merely directed to generating, collecting, and analyzing information.8 The court in Electric Power Group noted that “we have treated analyzing information by steps people go in their minds, or by mathematical algorithms, without more, as essentially mental processes within the abstract-idea category.”9 If this is the case, then the careful wording of AI patent applications will be increasingly important in getting those applications allowed. Indeed, the recent cases in the Federal Circuit and district courts with regards to AI have drawn attention to the treatment and wording of AI-related claims.
How have recent court decisions affected AI patentability?
More recent cases have drawn attention to the plight of AI in the courts. In 2018, in Finjan, Inc. v. Blue Coat Sys., Inc., the Federal Circuit found that claims directed to a method of virus scanning were found patent-eligible under Step 2A of the PTO’s subject matter eligibility test.10 Moreover, in a recent district court case pertaining to AI, PurePredictive, Inc. v. H2O.AI, Inc., the Northern District of California found that a claim to a machine learning predictive analysis framework was directed to “the abstract concept of the manipulation of mathematical functions and make[s] use of computers only as tools, rather than provid[ing] a specific improvement to computer related technology.”11 The Federal Circuit affirmed this decision without opinion in November 2018.12 These recent decisions regarding AI by both a district court and the Federal Circuit may present further limitations on the patentability of certain AI-related technologies.
How does the revised PTO guidance shape the outlook of AI patentability?
In January 2019, the PTO released revised subject matter eligibility guidance to reflect its interpretation of the Alice/Mayo test based on recent § 101 jurisprudence.13 This revision was carried out in part due to a “need for more clarity and predictability” in the application of the PTO’s subject matter eligibility test as well as “concern with the proper scope and application” of the “abstract idea” judicial exception.14 To address these concerns, the PTO modified Step 2A of the PTO’s Subject Matter Eligibility Guidance as incorporated into the Manual of Patent Examining Procedure (MPEP).15
The PTO revised Step 2A by (1) providing groupings of subject matter considered abstract ideas and (2) instructing that a claim is not directed to a judicial exception if the judicial exception is integrated into a practical application of that exception.16 Specifically, the first prong of Step 2A determines whether a claim is directed to a judicial exception (laws of nature, natural phenomena, and abstract ideas).17 The PTO limited abstract ideas to the following groupings: mathematical concepts, methods of organizing human activity, or mental processes.18 If the claim is directed to one of these judicial exceptions, the claim must be examined under the second prong of Step 2A, which considers whether the claim recites additional elements that integrate the exception into a practical application of that exception.19 If so, then the claim is patent eligible subject matter, but if not, the claim must be examined under Step 2B, which considers whether the claim is directed to an inventive concept.20
The revised § 101 guidance, in addition to the new § 112 guidance on examining computer-implemented invention, may be good news for potential applicants of AI patents. Since the § 101 guidance narrows the scope of the judicial exceptions by limiting abstract idea subject matter to specific groupings, AI technology may not fall into any of these categories, since many of the necessary actions cannot be practically performed or applied in the human mind due the amount of processor power required or data to be analyzed. For example, the PTO’s guidelines provide an example claim directed to a “computer-implemented method of training a neural network for facial detection” that it states does not recite any of the judicial exceptions because (1) it does not recite any mathematical relationships, functions, or calculations, (2) it does not recite a mental processes because the recited steps required to train a neutral network are “not practically performed in the human mind,” and (3) it does not recite any method of organizing human activity.21
Moreover, under the revised Step 2A, a claim may be considered “integrated into a practical exception” depending on the arguments and language used in the claim. For example, the PTO guidelines provides another sample claim directed to a “method for adaptive monitoring of traffic data through a network appliance connected between computing devices in a network.” The PTO guidelines state that this claim would be patentable under Step 2A because it is integrated into a practical application and the “claim as a whole is directed to a particular improvement in collecting traffic data” by “limit[ing] collection of additional … protocol data … which avoids excess traffic volume on the network … [and] the collected data can then be used to analyze the cause of the abnormal condition.”22 Thus, the patentability of AI-related inventions under the revised PTO guidelines may depend on the details of the claim. Those in the AI space may wish to monitor the developments in this area carefully to watch how the PTO and Federal Circuit interpret this new guidance.
In light of the revised PTO guidance and recent Federal Circuit jurisprudence, the distinction between an unpatentable application of an abstract idea and a patentable improvement to a technology seems to be a fine line defined by careful drafting. In an ideal world, a uniform interpretation of the patent statutes between the PTO examiners could ultimately generate a more efficient patent system. Clearer patent language could empower the patent community to shift away from the emphasis on the “draftsman’s art.” However, absent a shift in legal interpretation, the future of AI patents may depend for now on thoughtful patent drafting.
1 See, e.g., The Top IP Topics to Watch in 2019, Baker Botts (Dec. 19, 2018), http://www.bakerbotts.com/news/2018/12/the-top-ip-topics-to-watch.
2 See, e.g., Elec. Power Grp., LLC v. Alstom S.A., 830 F.3d 1350, 1354-56 (Fed. Cir. 2016).
3 U.S. Patent and Trademark Office announces revised guidance for determining subject matter eligibility, U.S. Patent and Trademark Office (Jan. 4, 2019), https://www.uspto.gov/about-us/news-updates/us-patent-and-trademark-office-announces-revised-guidance-determining-subject.
4 U.S. Patent and Trademark Office, Class 706, Data Processing - Artificial Intelligence, https://www.uspto.gov/web/patents/classification/uspc706/defs706.htm (last visited Jan. 24, 2019).
6 See, e.g., Aaron Gin et al., A Look at the Patenting Trends for Artificial Intelligence, Law360 (Nov. 20, 2018), https://www.law360.com/articles/1103089/a-look-at-the-patenting-trends-for-artificial-intelligence.
7 Elec. Power Grp., 830 F.3d at 1350.
8 Id. at 1354-56.
9 Id. at 1354.
10 Finjan, Inc. v. Blue Coat Sys., Inc., 879 F.3d 1299, 1305-6 (Fed. Cir. 2018).
11 PurePredictive, Inc. v. H2O.AI, Inc., No. 17-cv-03049-WHO, 2017 WL 3721480, at *1 (N.D. Cal. Aug. 29, 2017).
12 PurePredictive, Inc. v. H2O.AI, Inc., 741 Fed.Appx. 802 (Mem) (Fed. Cir. Nov. 7, 2018).
13 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (Jan. 7, 2019).
14 Id. at 50.
17 Id. at 52.
19 Id. at 54.
20 Id. at 56.
21 Claim: A computer implemented method of training a neural network for facial detection comprising: collecting a set of digital facial images from a database; applying one of more transformations to each digital facial image including mirroring, rotating, smoothing, or contrast reduction to create a modified set of digital facial images; creating a first training set comprising the collected set of digital facial images, the modified set of digital facial images, and a set of digital non-facial images; training the neural network in a first stage using the first training set; creating a second training set for a second stage of training comprising the first training set and digital non-facial images that are incorrectly detected as facial images after the first stage of training; and training the neural network in a second stage using the second training set. U.S. Patent and Trademark Office, Subject Matter Eligibility Examples: Abstract Idea, at 8-9 (Jan. 7, 2019) https://www.uspto.gov/sites/default/files/documents/101_examples_37to42_20190107.pdf.
22 Claim: A method for adaptive monitoring of traffic data through a network appliance connected between computing devices in a network, the method comprising: collecting, by the network appliance, traffic data relating to the network traffic passing through the network appliance, the traffic data comprising at least one of network delay, packet loss, or jitter; comparing, by the network appliance, at least one of the collected traffic data to a predefined threshold; and collecting additional traffic data relating to the network traffic when the collected traffic data is greater than the predefined threshold, the additional traffic data comprising Netflow protocol data. Id. at 10-11.
ABOUT BAKER BOTTS L.L.P.
Baker Botts is an international law firm whose lawyers practice throughout a network of offices around the globe. Based on our experience and knowledge of our clients' industries, we are recognized as a leading firm in the energy, technology and life sciences sectors. Since 1840, we have provided creative and effective legal solutions for our clients while demonstrating an unrelenting commitment to excellence. For more information, please visit bakerbotts.com.