As has been discussed here previously, the current position of the USPTO is that Artificial Intelligence (AI) cannot be an inventor. The USPTO’s position on the matter has been challenged in district court in Thaler v. Iancu, et al, 1:20-cv-00903. Although the case is not yet complete, it is likely that the court will side with the USPTO and hold that AI cannot be considered to be an inventor of a US patent application based on 35 U.S.C. § 100(f), which clearly states that “the term ‘inventor’ means the individual, or, if a joint invention, the individuals collectively who invented or discovered the subject matter of the invention.” Although Thaler, the Plaintiff in the case, has made a number of policy arguments for inclusion of AI inventorship, a ruling in his favor would be a surprise.
One of the arguments put forward by Thaler, in defense of AI inventorship, is that it would be a travesty of justice if an invention made by AI would be entirely excluded from patent protection just because the inventor is not a natural person. This argument assumes that if a natural person were to take credit for an invention made by AI, this also would result in an invalid patent due to improper naming of the true inventor. But would this be the case? Is it possible that an AI-conceived invention would, in fact, have an inventor who is a natural person?
In order to answer this question, it would be worthwhile to re-visit the law defining inventorship. MPEP §2109 does a good job of stating the law on this issue: “The definition for inventorship can be simply stated: ’The threshold question in determining inventorship is who conceived the invention. Unless a person contributes to the conception of the invention, he is not an inventor. … Insofar as defining an inventor is concerned, reduction to practice, per se, is irrelevant [except for simultaneous conception and reduction to practice, Fiers v. Revel, 984 F.2d 1164, 1168, 25 USPQ2d 1601, 1604-05 (Fed. Cir. 1993)]. One must contribute to the conception to be an inventor.’ In re Hardee, 223 USPQ 1122, 1123 (Comm’r Pat. 1984). ). A person who shares in the conception of a claimed invention is a joint inventor of that invention. In re VerHoef, 888 F.3d 1362, 1366-67, 126 F.2d 1561, 1564-65 (Fed. Cir. 2018). See also Board of Education ex rel. Board of Trustees of Florida State Univ. v. American Bioscience Inc., 333 F.3d 1330, 1340, 67 USPQ2d 1252, 1259 (Fed. Cir. 2003) (‘Invention requires conception).’”
But what is the definition of conception? One of the best definitions is found in Bosies v. Benedict, 27 F.3d 539, 543, 30 USPQ2d 1862, 1865 (Fed. Cir. 1994), which held that in order to establish “conception,” the inventor must form a definite and permanent idea of the complete and operable invention. At first blush it seems difficult to imagine that any natural person could be considered to “form a definite and permanent idea of the complete and operable invention” when AI develops an invention. However, any AI that exists today or is likely to exist in the future will have a natural person that initiates the AI and sets it upon a task. Although this person may not be fully aware, at the time of bringing the AI online, what specifically may be discovered by the AI, any discovery developed by the AI in some ways can be considered an extension of this initial step. Therefore, this natural person, although at the point of bringing the AI online cannot be considered an inventor, may nevertheless become an inventor at the point that the AI has developed a complete and operable invention. At this stage, the person’s initial act of putting the AI into motion can be considered to be complete. As a result, it can reasonably be considered that the person initializing the AI has conceived the invention and is thus the true inventor.It remains to be seen how the Thaler v. Iancu case will eventually resolve. However, it is reasonable to assume that this will not be the last case in which AI plays a major role in the conception of the invention.
 See Thaler v. Iancu, et al, 1:20-cv-00903