AI & Patentability

Artificial Intelligence plays a significant role in today’s world. In fact, if you are a modern era human, you cannot live without it.

Artificial intelligence (AI) has many uses and applications, including speech recognition, customer service, and innovation for improving society. Some people even utilize AI to help solve problems, invent new products, and develop new methods. For instance, AI can identify flaws in pharmaceutical products, create new eco-friendly compositions, and analyze biological samples in manufacturing. Consequently, some inventors choose to attribute the invention to AI when applying for patent protection.

Background & Definition

In 1950, Alan Turing, often called the “father of computer science,” initiated the AI conversation by posing, “Can machines think?” To address this question, Turing proposed the “Turing Test,” in which a human would be tasked with distinguishing between text responses produced by a human and those produced by a computer. While artificial intelligence has progressed beyond Turing’s original test, description, and comprehension, it still represents the beginning of an ongoing dialogue and discussion about AI.

In a simple definition, artificial intelligence, commonly called AI, is a branch of computer science that utilizes data sets and algorithms to identify patterns, learn from experience, and make predictions. Additionally, AI includes machine learning and deep learning, often mentioned together.

Types of AI

  • Weak AI, also known as Artificial Narrow Intelligence (ANI) or Narrow AI, is a type of artificial intelligence designed and trained to perform specific tasks. Most AI systems that the average person encounters today, such as Amazon’s Alexa or Apple’s Siri, are examples of Weak AI.
  • Artificial General Intelligence (AGI) and Artificial Super Intelligence (ASI) are two types of Strong AI. AGI would possess human-level intelligence and self-awareness to plan, learn, and solve problems. ASI, or super intelligence, would have intelligence surpassing human understanding. These types of AI remain theoretical and can only be found in science fiction.

Out of the two types mentioned earlier, the second one is more likely to be presented as an inventor. This is because the second type is usually more innovative and creative, which are essential qualities for inventors. Additionally, the second type tends to focus more on problem-solving and finding solutions, which are also crucial for inventors. Overall, while both types can succeed in many fields, the second type is often better suited for inventing and creating new things.

It may be helpful to consider the topic of AI as an inventor within the framework of U.S. jurisdiction.

A Brief History of Patents in the United States

One must understand the patent process and system within the United States to discuss the patentability of inventions where AI is listed as the inventor.

The Constitution of the United States lays the foundation for the governance of patent law through the Intellectual Property Clause. Congress has the power to promote the progress of science and useful arts by securing authors and inventors the exclusive right to their respective writings and discoveries.

Although the Intellectual Property Clause provides the initial framework for patent law within the United States, the Patent Act of 1790 (the 1790 Act) established the first patent laws in the US. These laws aimed to grant patents for up to fourteen years, thus promoting the progress of the useful arts for inventions considered sufficiently helpful and essential.

The Patent Act of 1793 replaced the 1790 Act and introduced the definition of patentable subject matter. This definition includes four categories:

  • Manufacturers,
  • Compositions of matter,
  • Processes,
  • and machines.

The patentable subject matter does not have laws of nature, abstract ideas, or natural phenomena. Within a patent application, the claim precisely describes the patentable subject matter for which protection is sought.

Currently, in the United States, three requirements must be met to obtain a patent for a new and useful invention:

  • The invention must be novel,
  • have utility,
  • and be non-obvious.

Additionally, a written description shall accompany all patent applications, providing specific and concise terms and any other additional specifications.

In the following section, a recent case where AI was listed as an inventor will be explained.

Thaler v. Hirshfeld

In Thaler v. Hirshfeld, the primary issue was surrounding inventorship and whether artificial intelligence adequately satisfied the necessary requirements for inventorship, such as satisfying the oath or declaration portion of the inventorship requirement.

The conception of an invention is the complete creation of the idea in the inventor’s mind, including its detailed practical application.

Before obtaining a patent for an invention, it must be reduced to practice. The reduction can be achieved through actual or constructive reduction to practice. Conception and reduction to practice are established when the invention is clear and can be reduced to practice by an individual skilled in that art without undue experimentation.

Stephen Thaler listed his artificial intelligence machine DABUS as the inventor on two patent applications filed with the USPTO on July 29, 2019. Thaler submitted a substitute statement on behalf of DABUS to fulfill the oath or declaration requirement, but the USPTO refused to process the applications.

Thaler sued Andrew Hirshfeld and the USPTO. Hirshfeld, as the Commissioner for Patents, was the defendant in this patent appeal case. In his complaint, Thaler claimed that rejecting his patent applications was “arbitrary” and “an abuse of discretion.” He requested that his applications be reinstated, the prior decision be vacated, and that patent applications for AI-generated inventions not be rejected due to the absence of a natural person inventor. Thaler requested that AI-generated inventions list artificial intelligence as the inventor on patent applications if the A.I. meets the necessary criteria for inventorship.

The USPTO argued that their interpretation of the Patent Act aligns with past and present interpretations of the statutory language. At the same time, Thaler contended that A.I. meeting criteria should be listed as an inventor if the USPTO did not consider alternative interpretations or exclude A.I. through evidence.

The court deemed Thaler’s argument invalid, citing the USPTO and Hirshfeld’s interpretation of statutes and prior case law. The court then addressed whether an inventor must be human, focusing on the term “individual.”

The court concluded that the word “individual” in the Patent Act refers to natural persons, meaning that inventors must be natural persons. The court also noted that artificial intelligence systems are not typically considered individuals.

The court ultimately stated that artificial intelligence cannot be listed as an inventor on a patent application and instead pointed to Congress for a change to be made, indicating otherwise, thus expanding the scope of patent law.

This is because it is not the court’s job to overcome the plain language of a statute for a policy concern; that is a matter to be handled by Congress. Finally, the court stated that if an “individual” were meant to include artificial intelligence, therefore allowing A.I. to satisfy the inventorship requirement of a patent application, Congress would have selected a different word when constructing the statutory language of the Patent Act. As a result, the court ultimately upheld the USPTO’s decision to refuse to process Thaler’s patent applications listing DABUS as the inventor.

Consequently, in the United States, A.I. is not recognized as an inventor by law. This decision was based on the interpretation of the term “individual,” which excludes machines.

For more information on the patentability of AI in the UK, check out episode 150 of the Law Pod UK podcast.


As mentioned in the present article, artificial intelligence (AI) is becoming increasingly important in the legal field. While there are still some limitations in recognizing AI as an inventor in the U.S. court system, it is clear that we are entering a new era and must adapt accordingly. As technology continues to advance, it is crucial that we stay informed and up-to-date on the latest developments in AI and its applications in the legal industry.

Karimi & Associates Law Firm, featuring a team of specialists in several areas including Patent and Intellectual Property, is prepared to provide full legal services in this field of law. In case you require legal assistance or consultancy, contact us to enjoy our top-tier services.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top