Artificial Intelligence (AI) – most notably generative AI like Chat GPT, Midjourney and Google's Veo 3 text-to-video model – is rapidly reshaping industries, revolutionizing everything from content creation to complex problem-solving. As AI-generated works become more prevalent, novel legal issues concerning ownership and protection of intellectual property (IP) rights have emerged. IP law covering patents, trade secrets, copyrights, and trademarks exists to safeguard innovations, artistic expressions, brand identities, and valuable secrets. However, AI's ability to generate content with varying levels of human contribution blurs the boundary lines relied upon by conventional IP law and challenges traditional rationales behind granting IP rights that reward and incentivize innovation. Understanding these evolving dynamics is essential for businesses and creators navigating the legal and regulatory changes to IP protections in the age of AI-generated content.
In Part One of this series, we explore how Patent Law is adapting to this new frontier and what it means for inventors looking to use AI to develop their inventions. Stay tuned for additional posts where we will examine how using AI can impact rights in trade secrets (Part Two), copyrights (Part Three) and trademarks (Part Four).
Patents – The Potential Perils of AI Inventorship
Patents protect inventions by granting the inventor exclusive rights to make, use or sell the patented innovation in a country for a certain period of time, typically 20 years. In exchange for these powerful rights, patent owners disclose to the public what they invented in sufficient detail to enable the public to practice the invention after it expires. Patents can protect new and useful processes, machines, compositions of matter, and improvements thereto, ensuring inventors can benefit from their creations while encouraging technological advancement.
The proposition of AI inventorship has already seen its first challenges. In Thaler v. Vidal, Stephen Thaler sought patent protection for inventions allegedly created solely by his AI system, DABUS, without naming a human inventor. The U.S. Court of Appeals for the Federal Circuit ruled that AI cannot be listed as an inventor or co-inventor on a patent application, reasoning that U.S. patent laws required human inventorship, and rejected the DABUS patent for improper inventorship. However, the Court was careful to explain that they did not answer the thornier issue of "whether inventions made by human beings with the assistance of AI are eligible for patent protection."
In early 2024, the U.S. Patent and Trademark Office (USPTO) published guidance on AI-assisted inventorship. While they caution that at least one human must make a "significant contribution" to each claimed invention, there is no categorical ban on AI-assisted inventions. While this is still a nascent issue, courts will likely consider the existing Pannu factors for inventorship, which determine inventorship by looking to whether each human inventor "(1) contribute[d] in some significant manner to the conception . . . of the invention, (2) [made] a contribution to the claimed invention that is not insignificant in quality, when that contribution is measured against the dimension of the full invention, and (3) [did] more than merely explain to the real inventors well-known concepts and/or the current state of the art." The USPTO's Guidance on AI Assisted Inventions counsels that human effort to "construct[] the prompt in view of a specific problem to elicit a particular solution" or "make[] a significant contribution to the output to create an invention" may qualify as a significant contribution, while "only presenting a problem to an AI system" or "merely recogniz[ing] and appreciat[ing] the output of an AI system as an invention" is likely insufficient.
Beyond complicating inventorship, the proliferation of AI tools further introduces the likelihood of potential misappropriation of others' IP. Because AI models are trained using massive datasets comprised of publicly available data that may be covered by existing IP, the output of an AI model needs to be closely scrutinized when integrated into creative or inventive works.
Key Takeaways
While patents require significant human involvement in conception of the invention, it is still possible to use AI as a tool to implement ideas and reduce them to practice. As AI tools become more integrated into regular workflow, it is important to consider how using AI can impact potential IP rights.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
