A.I. Not Yet an Inventor, but Maybe in the Future?
Recently, a Missouri physicist attempted to introduce a new player in the field of innovation: his self-invented artificial intelligence ("A.I.") tool. On July 29, 2019, Dr. Stephen Thaler filed two patent applications, therein listing the inventor as “DABUS.” He also listed the family name of the inventor as an “invention generated by artificial intelligence.”
In filing this application, Dr. Thaler tried to push the limits on who, or what, the U.S. Patent Office (“USPTO”) would accept as the primary inventor. Dr. Thaler’s brainchild, the Device for Autonomous Bootstrapping of Unified Sentience (“DABUS”), is a creative neural network. DABUS specifically combines multiple separate neural networks that all process random stimuli to generate potential ideas. DABUS then compares the different ideas the networks produce to curate the current idea, and it uses those comparisons to produce better results – a constant positive feedback loop for A.I. improvement. Dr. Thaler used this basic framework to argue that his machine is sentient and considered an inventor.
However, the United States Patent Office (“USPTO”) declined to agree with Dr. Thaler that an A.I. network can invent a patent-worthy widget. After the USPTO reviewed Dr. Thaler’s initial application, they issued him a “Notice to File Missing Parts of Non-Provisional Application.” In this notice, they cited a failure to list the primary inventors legal name, rejecting his idea that DABUS satisfied this role. Dr Thaler challenged his rejected application, filing suit in United States District Court for the Eastern District of Virginia, Thaler v. Hirshfeld, 558 F. Supp. 3d 238 (E.D. Va, 2021).
The Eastern District agreed with the USPTO, relying on the language of the Patent Act. The language of the Patent Act refers to an inventor as an “individual.” Using “plain language” interpretation, the Court found that the statutory term “individual” required a natural person – which DABUS could never become. Referring to a 2012 U.S. Supreme Court decision in Mohamad v. Palestinian Authority, 566 U.S. 449 (2012) that ruled the term “individual” within the Torture Victim Protection Act to mean a “natural person,” the Eastern District analogized that the inventor must also be a natural person. Bolstering this decision, many Federal Circuit Courts previously rejected a corporation as a qualifying inventor for the USPTO’s purposes.
In Beech Aircraft Corp. v. Edo Corp., 990 F.2d 1237 (Fed. Cir. 1993), the Federal Circuit held that a corporation “could never have been declared an ‘inventor,’ as [the corporation] was merely a corporate assignee and only natural persons can be ‘inventors’” (citing 35 U.S.C. §§ 115-118). Although not addressing artificial intelligence specifically (because it wasn’t invented yet) the precedent was clear that “inventors must be natural persons” and “only natural persons can be ‘inventors,’” thereby supporting the USPTO’s position. Dr. Thaler argued that the applicable statutes pre-dated artificial intelligence and therefore lacked consideration of modern technology.
Dr. Thaler further theorized that if “Congress had contemplated this artificial intelligence issue, it would have included artificial intelligence machines within the definition of ‘inventors.’” The Court and USPTO noted that Congress defined ‘inventor’ as an ‘individual’ through the America Invents Act in 2011, when artificial intelligence was already in existence.” See Pub. L. 112-29, § 3(a), 125 Stat. 285 (Sept. 16, 2011); see also H.R. Rep. No. 112-98 (June 1, 2011), available at 2011 U.S.C.C.A.N. 67, 67.
Thus, the Eastern District of Virginia held Dr. Thaler’s policy arguments premature but credited his concept by stating“[a]s technology evolves, there may come a time when artificial intelligence reaches a level of sophistication such that it might satisfy accepted meanings of inventorship. But that time has not yet arrived, and, if it does, it will be up to Congress to decide how, if at all, it wants to expand the scope of patent law.”
Dr. Thaler is appealing this ruling, and his only real argument is that the policy of considering DABUS a tool, rather than an individual, be reconsidered. Regardless of the appeal’s outcome, Dr. Thaler’s lawsuit is just a precursor to future, difficult questions courts are soon going to answer about the applications of artificial intelligence.