AI as a Tool for the "PHOSITA"

January 27, 2022

An ongoing discussion in the field of artificial intelligence (AI) and patent law relates to the implications of AI toward the standard for a person having ordinary skill in the art (“PHOSITA”). While the PHOSITA is a legal fiction, resolving the level of ordinary skill in the pertinent art is one of the Graham factors enunciated by the Supreme Court in the framework for determining obviousness under 35 U.S.C. §103.[1] 

However, this post is not about what the PHOSITA is for an invention in the field of AI itself, such as an invention related to neural networks.  Rather, it pertains to the fact that AI is a tool that is potentially applicable to a variety of industries and applications.  In other words, AI itself is tool that can enhance the productivity and output of persons having skills in many arts, and therefore AI could be assumed to be available to the PHOSITA when making a determination of obviousness even when the invention is not AI-related.

For instance, in October 2020, the USPTO published a report entitled “Public Views on Artificial Intelligence and Intellectual Property Policy,”[2] which takes a comprehensive look at a wide variety of stakeholder views on the impact of (AI) across the IP landscape.  The USPTO had posed the question “Does AI impact the level of a person of ordinary skill in the art?”[3] One of the responses from a commentator to this question stated “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.”[4]  Another commentator stated “Just as the existence of test tubes impacts the level of a person of ordinary skill in the chemical arts, and just as the existence of general purpose computers impacts the level of a person of ordinary skill in the software arts (and many others), so [too] would AI affect the level of skill in the arts where it can be made useful.”[5]  However, in the same report, another commentator cautioned against “declaring that all fields of innovation are now subject to the application of conventional AI.” 

For the time being, it is likely that this issue will only arise on a case-by-case basis without further guidance from the USPTO, especially in view of the flexibility provided by U.S. case law in defining a person of ordinary skill in the art.  Keep in mind the following statements from the Supreme Court.

"A person of ordinary skill in the art is also a person of ordinary creativity, not an automaton." KSR, 550 U.S. at 421, 82 USPQ2d at 1397. "[I]n many cases a person of ordinary skill will be able to fit the teachings of multiple patents together like pieces of a puzzle." Id. at 420, 82 USPQ2d at 1397. Office personnel may also take into account "the inferences and creative steps that a person of ordinary skill in the art would employ." Id. at 418, 82 USPQ2d at 1396.

Therefore, if it would only take “ordinary creativity” to use AI to accomplish a complex task, then at what point is it reasonable to assume AI could be used to fill in the gaps between what is in the prior art and in the claims of a patent application?  This is certainly not a solved legal issue, and patent practitioners should currently give extra scrutiny to an Examiner that attempts to use AI in this manner when making a rejection.

While it is entirely possible that the present legal framework for assessing the PHOSITA is adequate for determining how to treat AI as a tool in a variety of fields, the greater tension may exist in determining a standard for “conventional AI.”  For instance, while it is common knowledge that certain tasks are made much easier through AI (such as some complex forms of pattern recognition), how readily available would AI software need to be for the Examiner to assume that the PHOSITA would have access and understanding of how to use a “conventional” version of AI? 

We will monitor this issue with great interest going forward. 

[1] Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966).


[3] Id. at p. 11.

[4] Id. at p. 12.

[5] Id. at pp. 12-13.