Senior Counsel Matthew Horton and Associate Austin Kim published the article, “Inventorship: Why AI Is Not Smart Enough Yet,” in Managing IP. The article looks at the implications of a recent USPTO decision that said that artificial intelligence cannot be an inventor on a patent.
“The global patent community is currently engaged in a discourse on various policy issues surrounding patents for artificial intelligence (AI). One of the questions under discussion is whether an AI agent (as in, not a human) is, or should be, eligible to be an inventor on a patent application. In a decision published on April 22 2020, the USPTO responded with a categorical “no.” The decision rejected the listing of an AI agent – dubbed “Device for Autonomous Bootstrapping of Unified Sentience” (DABUS) – as the inventor, in a highly publicised patent application. The decision explained that an AI agent cannot meet certain statutory definitions for an inventor or the jurisprudential tests for determining inventorship under US law.”
“In the USPTO’s view, the Patent Act suggests an inventor must be a natural person because the statutory language describes an inventor with terms such as “whoever,” “himself or herself,” and “individual.” The act also describes inventor obligations in terms of actions that cannot be performed by any other entity besides human beings, such as requiring an inventor to “execute an oath or declaration.” In addition to the statutory language, the decision further relies on case law defining the contours of inventorship. Generally, inventorship turns on the question of “conception.” The legal tests for conception require a person to conceive of and contribute to the inventive subject matter in the claims. In case law, these tests expressly define conception as arising from “the mind of the inventor.” Therefore, according to the USPTO, only natural persons can satisfy the requirements for inventorship,” Horton and Kim wrote.
Read the full article here. (Subscription required)