Osha Attorneys
Osha Attorneys
Osha Attorneys
Osha Attorneys
Attorney search
Search by

The collective expertise of our global team distinguishes OBWB in the field of Intellectual Property Law. We align our best resources to meet each client's specific needs and we treat each matter with the highest degree of attention and care.

Summary of USPTO Report on Public Views on Patenting AI Inventions

By Silvia V. Rusnac

日本語

A key priority of the United States Patent and Trademark Office (USPTO) is to maintain United States leadership in innovation, especially in emerging technologies, including artificial intelligence (AI). To further this goal, on August 27, 2019, the USPTO issued a request for comments (RFC) on patenting AI inventions. The RFC sought feedback from stakeholders on a variety of patent policy issues, such as AI’s impact on inventorship and ownership, eligibility, disclosure, and the level of ordinary skill in the art. The comments regarding patenting of AI inventions are summarized below, organized by the RFC questions.

Part I—Responses to the RFC on Patenting Artificial Intelligence Inventions, issued on August 27, 2019

1.     What are elements of an AI invention? For example: The problem to be addressed (e.g., application of AI); the structure of the database on which the AI will be trained and will act; the training of the algorithm on the data; the algorithm itself; the results of the AI invention through an automated process; the policies/weights to be applied to the data that affects the outcome of the results; and/or other elements.

 This question sought to identify broadly the elements of an AI invention that may be subject to patentability. Four common answers arose: 

(1) The elements disclosed in the question constitute a non-exclusive list of elements of an AI invention. 

(2) AI can be understood as computer functionality that mimics cognitive functions associated with the human mind (e.g., the ability to learn). 

(3) AI inventions can be categorized as follows: (a) inventions that embody an advance in the field of AI; (b) inventions that apply AI (to a field other than AI); and (c) inventions that may be produced by AI itself. 

(4) Undue effort should not be expended on defining AI, which is dynamic and will be subject to fundamental change in the coming years.

2.     What are the different ways that a natural person can contribute to conception of an AI invention and be eligible to be a named inventor? For example: Designing the algorithm and/or weighting adaptations; structuring the data on which the algorithm runs; running the AI algorithm on the data and obtaining the results. 

The development of AI may present many opportunities for invention in the areas of designing an AI algorithm, implementing particular hardware to enhance an AI algorithm, or applying methods of preparing inputs to an AI algorithm. Several innovators may be involved in the development of an AI system. 

The vast majority of public commenters asserted that current inventorship law is equipped to handle inventorship of AI technologies. Many commenters suggested that assessment of conception should be fact-specific, as in the analysis done today.  However, some commenters were of the opinion that mere running of an AI algorithm on data and obtaining results is unlikely to qualify as a contribution to conception. 

3.     Do current patent laws and regulations regarding inventorship need to be revised to take into account inventions where an entity or entities other than a natural person contributed to the conception of an invention? 

Although AI has potential for autonomous creation, present AI technology appears to be within the realm of narrow, application-specific objectives. The majority of commenters viewed the concept of artificial general intelligence (AGI) — intelligence akin to that possessed by humankind and beyond—as merely a theoretical possibility that could arise in a distant future.

Conception is the formation in the mind of the inventor of a definite and permanent idea of the complete and operative invention. The Patent Statute, Title 35, U.S.C., is replete with language indicating that the inventor of a patent application must be a natural person. 

The majority of commenters indicated that there is no need for revising patent laws and regulations on inventorship to account for inventions in which an entity other than a natural person contributed to the conception of an invention. However, some commenters suggested that the USPTO should revisit the question when machines begin achieving AGI. 

4.     Should an entity or entities other than a natural person, or company to which a natural person assigns an invention, be able to own a patent on the AI invention? For example: Should a company who trains the artificial intelligence process that creates the invention be able to be an owner? 

The vast majority of commenters stated that no changes should be necessary to the current U.S. law—that only a natural person or a company, via assignment, should be considered the owner of a patent or an invention. However, a minority of responses stated that while inventorship and ownership rights should not be extended to machines, consideration should be given to expanding ownership to a natural person: (1) who trains an AI process, or (2) who owns/controls an AI system. 

5.     Are there any patent eligibility considerations unique to AI inventions? 

Many commenters asserted that there are no patent eligibility considerations unique to AI inventions. That is, AI inventions should not be treated any differently than other computer-implemented inventions. Claims to an AI invention that fall within one of the four statutory categories and are patent-eligible under the Alice/Mayo test will be patent subject matter eligible under 35 U.S.C. § 101. Although many AI inventions are at risk of being characterized as certain methods of organizing human activity, mental processes, or mathematical concepts, the complex algorithms that underpin AI inventions have the ability to yield technological improvements. In addition, claims directed to an abstract idea will still be patent-eligible if the additional claim elements, considered individually or as an ordered combination, amount to significantly more than the abstract idea so as to transform it into patent-eligible subject matter. 

6.     Are there any disclosure-related considerations unique to AI inventions? For example, under current practice, written description support for computer-implemented inventions generally require sufficient disclosure of an algorithm to perform a claimed function, such that a person of ordinary skill in the art can reasonably conclude that the inventor had possession of the claimed invention. Does there need to be a change in the level of detail an applicant must provide in order to comply with the written description requirement, particularly for deep learning systems that may have a large number of hidden layers with weights that evolve during the learning/training process without human intervention or knowledge? 

35 U.S.C. § 112(a) has three separate and distinct disclosure requirements: written description, enablement, and best mode. These requirements apply to all applications examined before the USPTO, including those directed to AI inventions. The majority of commenters indicated that there are no unique disclosure considerations for AI inventions. However, some commenters indicated that there are significant and unique challenges to satisfying the disclosure requirements for an AI invention because the logic between a provided input and a generated output is in some respects unknown. 

7.     How can patent applications for AI inventions best comply with the enablement requirement, particularly given the degree of unpredictability of certain AI systems?

The comments suggest that there is no consensus on the predictability of AI systems.  While one commenter stated that most current AI systems behave in a predictable manner and that predictability is often the basis for the commercial value of practical applications of these technologies, another commenter noted that some AI inventions may operate in a black box because there is an inherent randomness in AI algorithms. Some commenters stressed that it may be difficult to enable (i.e., teach the public to make and use) certain AI inventions, as required by 35 U.S.C. § 112(a), and offered the topic for further exploration by the USPTO.

8.     Does AI impact the level of a person of ordinary skill in the art? If so, how? For example: Should assessment of the level of ordinary skill in the art reflect the capability possessed by AI?

Many commenters asserted that AI has the potential to affect the skill level of the hypothetical “ordinary skilled artisan,” thereby affecting the bar for nonobviousness. Numerous commenters suggested that the present legal framework for assessing the person of ordinary skill in the art is adequate to determine the impact of AI-based tools in a given field.  Some commenters elaborated that the level of skill in any art has traditionally grown over time based on the introduction of new technologies, and that once conventional AI systems become widely available such accessibility would be expected to enhance the abilities of a person of ordinary skill in an art. 

However, some commenters cautioned that such wide prevalence of AI systems has not yet permeated all fields and counseled against declaring that all fields of innovation are now subject to the application of “conventional AI.” Others interpreted the question to assume a future state in which AGI exists and machines have intelligence comparable to humans or beyond. Those interpreting the question in this manner suggested that such machines are not persons and, therefore, would not affect the legal standard of a “person” of ordinary skill in the art. 

9.     Are there any prior art considerations unique to AI inventions? 

The majority of commenters stated that there were no prior art considerations unique to AI inventions—that current standards were sufficient. A minority of commenters indicated that there were prior art considerations unique to AI inventions, many of which focused on the proliferation of prior art, such as the generation of prior art by AI, and the difficulty in finding prior art, such as source code related to AI. A minority of commenters indicated that while no prior art considerations unique to AI inventions currently existed, depending on how sophisticated AI becomes in the future, unique AI prior art could become relevant. 

10.  Are there any new forms of intellectual property protections that are needed for AI inventions, such as data protection? 

Access to data for initial development and ongoing training is necessary for AI development. This means that data and datasets, including their collection and compiling, have value, particularly “big data" that may be analyzed computationally to reveal patterns, trends, and associations. Data protection under current U.S. law is limited in scope (e.g., under trade secret law), and the U.S. does not currently have intellectual property (IP) rights protections solely focused on data for AI algorithms. 

Commenters were nearly equally divided between the view that new IP rights were necessary to address AI inventions and the belief that the current U.S. IP framework was adequate to address AI inventions. Generally, however, commenters who did not see the need for new forms of IP rights suggested that developments in AI technology should be monitored to ensure needs were keeping pace with AI technology developments. Commenters did not provide concrete proposals on how any potential newly created IP rights should function, and many, from both sides of the divide, called on the USPTO to further consult the public on the issue. 

11.  Are there any other issues pertinent to patenting AI inventions that we should examine?

The commenters stressed the importance for examiner technical training and a call for memorializing guidance specific to AI for patent examiners.  Examples of suggestions provided in the comments are inviting industry trade groups to adopt formal recommendations for patent applications, stressing the importance of an “open ecosystem of research” to U.S. economic and scientific leadership in the field of AI, considering whether, given the dynamic nature of AI systems, the claims to such inventions need to be constantly updated as well, and considering issues outside of patentability, such as patent infringement and patent enforcement. 

12.  Are there any relevant policies or practices from other major patent agencies that may help inform USPTO’s policies and practices regarding patenting of AI inventions?

The USPTO participates in numerous global activities. While some commenters sought to have the USPTO continue its multilateral engagements on AI through World Intellectual Property Organization (WIPO) and other organizations, one commenter cautioned against further attempts to harmonize patent laws and procedures, especially as it relates to patenting AI, because U.S. patent law has long been the “gold standard” for patent protection and a major driver in the success of the U.S. innovation economy. 

Future Work

On October 30, 2019, the USPTO issued a second RFC related to the impact of AI on other IP policy areas, including copyrights, trademarks, database protections, and trade secret law. A summary of the comments regarding non-patent IP protection for AI innovation will be provided in a future article.  

The USPTO stated that it will use the USPTO report to focus issues for continued exploration of other measures it may take to bolster the understanding and reliability of IP rights for emerging technologies, such as AI. These steps may include further engagement with the public, additional guidance for stakeholders, and continued training for examiners on emerging technologies.  OBWB will provide further reports on USPTO guidelines as new guidelines becomes available.