You are currently viewing AI and Invention: A Legal Perspective

AI and Invention: A Legal Perspective

Artificial Intelligence (AI) has rapidly transformed the landscape of invention and innovation. Over the past few decades, the sophistication of AI systems has grown tremendously, enabling them to assist, and sometimes even surpass, human beings in many intellectual tasks. In the realm of invention, AI has been leveraged to create new processes, solve complex technical problems, and even generate inventions autonomously. This advancement has blurred the lines of intellectual property rights, particularly in the field of patent law. The current patent system, primarily designed with human inventors in mind, now faces new challenges associated with AI-generated inventions. The legal implications of this issue have sparked extensive research and debate, as legislators, inventors, and AI experts grapple with the question of how to adapt patent law for an increasingly AI-driven world.

AI systems have become integral in the process of creating new inventions. Through sophisticated machine learning algorithms and data analysis, these systems have the ability to identify patterns, generate ideas, and even refine these ideas into practical, tangible inventions. The role of AI is particularly prominent in fields like drug discovery, where it can analyze vast amounts of data more quickly and accurately than a human could. These capabilities not only increase the speed and efficiency of invention, but also open the door to new possibilities previously unattainable by human inventors alone. Nevertheless, the increasing autonomy of AI systems in the invention process necessitates a careful reevaluation and likely modification of our current intellectual property framework.

Understanding the Patent System

Patents represent a legal instrument that grants inventors exclusive rights to their inventions, thereby protecting them from exploitation by others. The primary purpose of a patent is to foster innovation by providing inventors with a period of exclusive commercial control over their inventions. This control is granted in exchange for the full disclosure of the invention, thereby promoting the advancement of knowledge and technology. By protecting the rights of inventors, patents serve as a crucial catalyst for technological progress, investment in research and development, and economic growth.

Brief Overview of the Current Patent Law

Under the current patent law, primarily governed by the Patent Act of 1952 in the United States, patents are granted to “whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof.” Essentially, to be deemed patentable, an invention should not only be novel, but it should also be non-obvious and have a practical application. Historically, the role of the “inventor” has been reserved for natural persons, typically interpreted as human beings. The patent application process involves the submission of a detailed description of the invention, which must demonstrate that a definite and permanent idea of the complete and operative invention existed in the mind of the inventor at the time of application. The patent, once granted, provides the inventor with the exclusive right to use, manufacture, and sell the invention in the country where the patent was issued, typically for a period of 20 years from the filing date. The current laws, however, do not explicitly address the eligibility or rights of non-human inventors such as AI systems, leading to complex legal and ethical questions in our rapidly evolving technological landscape.

Current Legal Status of AI Inventions

The legal status of AI-generated inventions continues to be a contentious issue under existing patent laws. These laws, as currently interpreted, do not recognize AI systems as inventors, given the long-standing requirement for inventors to be “natural persons.” The United States Patent and Trademark Office (USPTO) and other patent offices around the globe have generally maintained this position, rejecting patent applications that list AI systems or algorithms as the inventor. This stance was underscored in the recent decisions of the USPTO, the European Patent Office (EPO), and the UK Intellectual Property Office (IPO), all of which rejected patent applications for inventions created by an AI system known as “DABUS,” asserting that the current laws and conventions necessitate an inventor to be a human being. The issue is far from settled, however, with ongoing court cases and debates questioning whether the existing legal framework is adequate for addressing AI-generated inventions. As AI systems become increasingly sophisticated and autonomous, the need for new regulations, or the reinterpretation of existing ones, will continue to gain urgency.

Requirement for Human Inventors

One of the primary limitations in patenting AI inventions is the requirement for human inventors. Current patent laws, which were constructed in a time before the rise of AI, invariably emphasize the role of ‘natural persons’ in the creation of an invention. This requisite plays a pivotal role in defining who or what can be an inventor, hence directly influencing patent eligibility. Under this framework, an invention is typically considered as an embodiment of human ingenuity, an intellectual output resulting from human conception and execution. Consequently, the credit, as well as the legal protection offered by patents, is traditionally granted to human inventors. AI systems, despite their capability to generate innovative solutions, do not meet this criterion and are therefore excluded from being recognized as inventors. This requirement presents a significant challenge as we navigate into an era where AI is not merely an instrument of human inventors, but potentially an originator of novel and useful inventions in its own right.

The Issue of Autonomous Bootstrapping

Another crucial issue when discussing AI as inventors is the concept of autonomous bootstrapping. This refers to the ability of AI systems to iteratively improve and innovate without explicit human intervention, based on their own generated output. This is particularly significant in the context of machine learning, where AI systems utilize their own outputs as a basis for subsequent iterations, leading to progressively refined and more innovative solutions. Under the current legal framework, these self-generated and improved inventions would not be eligible for patent protection, since they lack an identifiable human inventor. This absence of legal protection raises concerns about the commercial viability of such inventions, potentially stunting investment and development in AI-driven research and innovation. Addressing this issue will require a fundamental rethinking of the traditional human-centric paradigm of invention, pushing for a more inclusive understanding that can accommodate the unique capabilities and potential of AI systems.

Case Studies and Legal Precedents

There have been several key cases and decisions related to the patentability of AI-generated inventions that have shaped the current legal understanding of this issue. One seminal case involves an AI system known as “DABUS.” In 2020, the USPTO rejected two patent applications that listed DABUS as the inventor, stating that under current law, only natural persons can be inventors. Similarly, the European Patent Office (EPO) and the UK Intellectual Property Office (IPO) also dismissed applications listing DABUS as the inventor, underscoring the human-inventor requirement in patent law.

Another noteworthy case is Thaler v. Commissioner of Patents in Australia, where the Federal Court of Australia recognized DABUS as an inventor. This marked a significant departure from previous rulings and opened the door to the possibility of AI systems being recognized as inventors in the future.

These cases highlight the ongoing legal ambiguity and international divergence surrounding the patentability of AI-generated inventions. They underscore the need for a comprehensive review and potential reformulation of existing laws and regulations to effectively address the evolving realities of AI innovation.

The varying decisions in different parts of the world have critical implications for AI inventions. Firstly, they create legal uncertainty for inventors and businesses relying on AI technology. Without clear guidelines on patentability, the protection of AI inventions remains ambiguous, potentially discouraging innovation and investment in AI research. Secondly, recognizing AI as possible inventors, as seen in the Australian case, could fundamentally alter the conception of invention and intellectual property, necessitating considerable adjustments in patent laws globally. Lastly, these divergent international views could lead to inconsistent patent rights across different jurisdictions, complicating the enforcement of patent protection for AI inventions. Overall, these decisions reflect an urgent need for international harmonization of laws to address the evolving landscape of AI innovation.


The challenges presented by AI-generated inventions in the realm of patent law are complex yet fascinating. The current legal framework, grounded in a human-centric concept of invention, is increasingly being challenged by the autonomous and innovative capabilities of AI systems. This elicits important questions about the adequacy of existing laws and the need for innovative legal thinking. While these challenges are daunting, they also present distinct opportunities. The recognition of AI systems as potential inventors could catalyze a significant shift in our understanding of invention and intellectual property. This could pave the way for more comprehensive protections for AI inventions, fostering an environment conducive to innovation and technological advancement. Furthermore, the ongoing international debates on this issue underscore the need for global cooperation and harmonization of laws, offering a valuable opportunity to build a more unified and effective international patent system that reflects the realities of the AI era.

Looking to Patent and Invention?

Please contact Arlen Olsen at Schmeiser, Olsen & Watts LLP at

About the Author

Mr. Olsen, a former adjunct professor of intellectual property law, has over 25 years of experience in all aspects of intellectual property law. Mr. Olsen is a founding Partner of Schmeiser, Olsen & Watts LLP and a former United States Patent Examiner. Mr. Olsen has prosecuted numerous patents that have been litigated and received damages in excess of $60 million dollars. Additional activities include teaching seminars and appearing as a guest lecturer on intellectual property matters for corporations and educational institutions and evaluating and consulting with clients regarding the scope, enforcement and protection of intellectual property rights.