The US Patent and Trademark Office (USPTO) recently issued what it labeled as Inventorship Guidance for AI-Assisted Inventions [Docket No. PTO-P-2023-0043]. Despite its name, the document provides little in the way of certainty that one could not garner from reviewing recent precedent addressing the issue of artificial intelligence (AI) inventions. To begin with, the USPTO warns that its guidance does not constitute substantive rulemaking and does not have the force and effect of law. Rather, [t]he guidance sets out agency policy with respect to the USPTO’s interpretation of the inventorship requirements of the Patent Act in view of controlling jurisprudence, but [r]ejections will continue to be based on the substantive law, and it is those rejections that are appealable to the PTAB and the courts. Adding to the confusion attendant to the actual purpose thereof, the guidelines admonish that, [t]o the extent that earlier guidance from the USPTO, including certain sections of the Manual of Patent Examining Procedure is inconsistent with the guidance set forth in such guidelines, USPTO personnel are to follow these guidelines, and [t]he MPEP will be updated in due course.
Putting aside whether this type of instruction to the patent examining corps constitutes rulemaking, it is clear that the USPTO intends these guidelines to be more than mere travelogue through recent decisions on the proper role of AI in patentable inventions. That said, it is worth noting the caselaw that the USPTO elected to address in the context of the policy underlying the US patent system, which is designed to encourage human ingenuity. The following is a synopsis of the more salient caselaw discussed in the guidelines:
In light of the foregoing (and other) decisions, the USPTO recognized that AI-assisted inventions are patentable where a natural person has made a significant contribution to the claimed invention, therefore qualifying as an inventor of such invention. As to the decision in Pannu regarding such significance, the USPTO acknowledged that [d]etermining whether a natural person’s contribution in AI-assisted inventions is significant may be difficult to ascertain, and there is no bright-line test. This recognition brought the USPTO to the specific guidelines it promulgated to help inform the application of the Pannu factors in AI-assisted inventions:
The USPTO also pointed to the extensive jurisprudence regarding conception in the context of inventorship disputes under pre-America Invents Act section 102(g), which the USPTO suggests may be instructive as to the significant nature of a purported inventor’s contribution. Notwithstanding that jurisprudence, the actual guidelines propounded raise more questions than they answer. For example, it is difficult to discern where an unpatentable intellectual domination of the fifth guideline ends and the specific prompt to elicit a particular solution from the second guideline begins.
Despite the questions they raise, the AI Guidance makes clear that the patentability of AI-assisted inventions will very much be a case-by-case inquiry, and it behooves a patent applicant to emphasize the role of a human inventor in the conception of the claimed invention. This can be done, for example, by fashioning claims to the extent possible to highlight the conceptual contribution of the human inventor. However, the USPTO added a cautionary note to the guidelines reminding applicants of the duties of disclosure and inquiry owed by patent applicants. Thus, the claims as drafted, and indeed the disclosure generally, must have a tether to the actual inventive work underlying the subject application.
Relatedly, the USPTO reiterated the requirement of an inventor’s oath accompanying every patent application. That oath is to include a representation that the inventor believes himself or herself to be the original inventor or an original joint inventor of a claimed invention in the application. As such oath is subject to criminal penalties for false statements contained therein, it is imperative that careful attention be paid to the actual inventive contribution of the inventor submitting the oath vis-à-vis the contribution of an AI system.
In sum, in seeking patent protection for AI-assisted inventions in light of the USPTO’s new guidelines, an applicant should consult with competent prosecution counsel capable of fashioning disclosures and associated claims designed to highlight the inventive contribution of human inventors consistent with the legal obligations governing the submission of information to the USPTO.
