Clicky chatsimple

AI Is Currently Unable To Obtain Patents For US Inventions

Category :

Workflow Automation

Posted On :

Share This :

Divergent opinions on whether AI may be recognized as an inventor globally.
Not just the United States of America forbids AI from being an inventor. Additionally, a general appeal court in the UK recently rejected Thaler’s patents’ AI inventorship. The United Kingdom’s patent statute, which stipulates that a “person” must be an “inventor” to be an actual deviser, was also examined by the Court of Appeal (Civil Division) of the United Kingdom.

However, because Australian law did not preclude such an interpretation, the Full Court of the Federal Court—a general court whose appeals can only be heard at the High Court—found that AI could be an “inventor” in Australia and acknowledged “the evolving nature of patentable inventions and their creators.” Thaler’s patent identifying AI as the inventor has been awarded by the South African Patent Office, which lacks a meaningful method for examining patent applications.

Consequences Of This Choice

What follows then? For the time being, Thaler has asked the Federal Circuit to rehear the matter. (Thaler v. Vidal, No. 21-2347, Paper No. 62 (Fed. Cir. Aug. 5, 2022)). In support of his request, Thaler claims that: (1) the dictionary definition of “individual” includes artificial intelligence (AI); (2) rejecting patent protection for AI-generated inventions is inconsistent with Supreme Court precedent that forbids rejecting patents based on the process of invention; and (3) the ruling deviates from the Supreme Court’s statutory interpretation approach in light of the rapid advancement of technology. (Id. at 10–11).
The USPTO’s decision to reject patent applications when the inventor is declared to be artificial intelligence (AI) is supported by the Federal Circuit’s notional bright line rule, which stipulates that a “inventor” must be a human. This rule offers clarity up front. The Court was able to decide a legal matter based on the record on appeal. This is as a result of the applicant’s AI designation. The more complicated topic of how much dependence on AI during the invention process would prevent a human from claiming inventorship was never addressed by the Court.
To put it another way, if artificial intelligence (AI) is employed as a tool for invention, then, similar to conventional inventions, the use of tools in the process of an invention does not prevent an applicant from claiming inventorship. However, the innovation is not eligible for patent protection if AI plays an innovative function equivalent to that of an inventor, according to a recent Federal Circuit opinion. This is a factual inquiry that may involve some extremely challenging line drawing, which the Thaler case managed to avoid.
Keep in mind that there is precedent for this kind of line drawing. It is a common requirement of our analysis of inventorship under U.S. law to determine whether an inventor is made or not by the contributions of human partners. If this is the case, the patent application lists the people as inventors and the patent prosecution proceeds. However, according to Thaler, the AI claims cannot be patented if the same inventorship examination determines that AI is the inventor.
Because utility and design patent applications are subject to the same inventorship statute, drafters of design patents may encounter comparable problems. Similar to utilitarian inventions, computer programs like CAD have been widely employed by designers as tools to help bring their ideas to reality. However, AI is being utilized more and more to optimize current designs (for several reasons, but one of the main goals is to minimize the amount of material required to create a product).
The AI could be perceived as being a co-designer when these modifications have an impact on how the design looks. We might soon come to a point where AI creates the entire design and humans only contribute to the final decision.
Similar problems with copyright law occur. The Copyright Office determined earlier this year that artificial intelligence (AI)-generated art could not be granted a copyright due to the absence of “human authorship.” Remarkably, Thaler was also involved in this case.) The Copyright Office declined to “depart from a century of copyright jurisprudence” on this issue, despite Thaler’s claims that this ruling was unconstitutional. This refusal included a 2018 ruling that determined a monkey could not file a lawsuit for copyright infringement over a photo it took after robbing a photographer of their camera.

Potential Courses Of Action

There are various IP protection measures that innovators might employ to overcome this possible barrier to patentability in AI. Drafting a utility patent specification about AI-related technology requires that the disclosure explicitly states that the technology is an instrument that humans employ to carry out the invention. As the gold standard for inventorship analysis, patent application claims should be written so that every component points to a human contribution.
Other types of IP protection are available if disclosing and claiming the invention in a way that limits AI to a tool and emphasizes human creativity is not feasible or desirable. Trade secret protection is one choice. Inventors are exempt from the disclosure and claim requirements of the patent statute due to the requirement to protect trade secret status by prohibiting public disclosure. When it comes to the difficulties in fulfilling those responsibilities brought about by AI inventions, this can be especially advantageous.
Moreover, trade secret protection does not require an inventorship determination. Therefore, whether an innovation comes from human or non-human inputs has no bearing on whether it is eligible for trade secret protection. Consequently, the Thaler threshold does not apply to trade secret protection because there is no legislative structure specific to patent protection, which gave rise to the Thaler test.
If a design satisfies the non-functionality condition and acquires secondary significance, it might still be viable to register for trade dress rights in the design. One kind of trademark that does not need the identity of a human designer is trade dress.
This case may serve as inspiration for Congress to review the Patent Act (and possibly the Copyright Act) in light of developing technology, while we wait for additional guidance from the Supreme Court. There are difficulties with this work. The necessity for an invention to declare that they believe they are the original inventor would also need to be changed by Congress.
Additionally, if the textual description requirement is designed to be understood by a machine in addition to or instead of a human, it may vary. However, granting AI inventorship might prevent a bias in favor of trade secret protection for AI, which would needlessly limit the amount of knowledge made available to the public. Furthermore, there might be a “bubble” of AI-generated innovations that are no longer patentable if Congress subsequently amended the statute (for example, because of the on-sale bar).
However, the question of whether AI needs the protection of inventorship runs against to all of these factors. AI inventions might be relatively inexpensive to produce, which would reduce the motivation to invest that comes with the possibility of obtaining patent rights. However, these innovations can be very lucrative in the future, making rivals susceptible to stealing them.
Lastly, there is still more confusion around these rights due to the issue of machine property rights. With all of the facets of this complex problem, it appears that Thaler may be the first case in a long line.