Exploring Patent Rights in the Realm of AI Development
In the rapidly evolving landscape of artificial intelligence (AI), jurisdictions worldwide are adapting their approaches to accommodate AI-related innovations and address the unique challenges they present.
Patent law, a critical mechanism for fostering innovation, is no exception. The patentability of AI innovations hinges on several core criteria: novelty, non-obviousness, and utility. However, as AI systems evolve, determining who holds rights to patents becomes more intricate, raising questions about the inventorship of AI-generated inventions.
Current patent law mandates human inventorship, a requirement that becomes problematic when AI systems contribute significantly to the invention process. Reforms in patent law are necessary to accommodate AI's role, with potential reforms including establishing distinct protocols for assessing AI inventions and refining definitions of inventorship and ownership.
The future of patent law in an AI-dominated world is poised to adapt to the rapid advancements in technology. International cooperation is essential in shaping patent law for AI, with harmonized regulations creating a cohesive patent landscape. The Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) sets minimum patent protection standards globally, including for AI technologies.
Landmark legal cases, such as IBM's Watson and Thaler v. Hirshfeld, demonstrate the complexities surrounding patent law and artificial intelligence. Courts continue to hold that AI patents must demonstrate concrete, technical innovations beyond merely applying AI or machine learning to known problems; abstract ideas or routine automation generally fail eligibility under 35 U.S.C. § 101.
Recent U.S. Patent and Trademark Office (USPTO) guidance aims to reduce overbroad rejections by clarifying examiners must look for technical improvements in AI inventions rather than generic applications of AI.
Potential reforms under discussion to address these challenges include redefining "inventor" to possibly include AI systems or accommodate AI contributions, creating new legal categories such as “AI-assisted inventions,” establishing frameworks for ownership and liability of AI-generated inventions, harmonizing patent eligibility and disclosure standards internationally, and modernizing patent examination processes with technology.
Japan’s recent court ruling rejecting AI as a patent inventor and the subsequent strategic IP program illustrate active legal-political engagement on these reforms, emphasizing the need for prompt legislative responses to keep pace with AI innovation. Meanwhile, the USPTO’s 2025 memo clarifies examination standards for AI patents but does not yet recognize AI as an inventor or create new patent categories.
In summary, the challenges revolve around defining inventorship, meeting disclosure and eligibility standards for AI’s unique output, and harmonizing laws internationally. Potential reforms focus on updating legal definitions and examination frameworks to reflect AI’s role while preserving the integrity and incentives of the patent system. These efforts are actively underway but remain contentious and evolving globally in 2025.
The Patent Cooperation Treaty (PCT) enables applicants to file a single patent application that is recognized in multiple jurisdictions, benefiting AI developers seeking to protect innovations across borders. International frameworks for patent offices could streamline the patenting process for AI technologies, ensuring consistent standards and promoting innovation.
Google's patent for machine learning algorithms optimizing ads in real-time has reshaped digital marketing strategies, allowing companies to reach targeted audiences with increased efficiency and effectiveness. However, challenges often arise in demonstrating the novelty, non-obviousness, and utility of AI systems, particularly in distinguishing between traditional invention processes and AI-generated outputs.
The patentability criteria for AI-related inventions often require reassessment, as traditional patent systems prioritize human inventorship but AI systems may lead to breakthroughs that do not fit conventional categorization. As the world continues to grapple with these issues, it is clear that the future of patent law will be shaped by the ongoing dialogue between lawmakers, technologists, and the global community.
