Artificial Intelligence (AI) is rapidly transforming the legal and technological landscapes — and patent law is no exception. As AI tools become integral to scientific discovery and product innovation, they are reshaping what qualifies as a patentable invention. The intersection of AI and patent law raises critical questions: Can AI-generated inventions be patented? How does AI impact the standard of what is “obvious”? Will AI help or hinder inventors in obtaining strong patents?
This article delves into how artificial intelligence will naturally affect patentability, exploring the legal, technical, and philosophical dimensions of this evolving relationship.
The Legal Foundation of Patentability
Patents are legal rights granted to inventors in exchange for public disclosure of new, useful, and non-obvious inventions. Under U.S. law, three main criteria determine whether an invention is patentable:
- Novelty – The invention must be new and not previously disclosed.
- Non-obviousness – The invention must not be an obvious development to someone skilled in the field.
- Enablement – The patent application must describe the invention clearly enough that others skilled in the art can reproduce it.
AI challenges all three of these pillars—particularly non-obviousness and enablement.
Who Can Be an Inventor? The DABUS Debate
A central issue in the AI-patent conversation is whether AI systems can be inventors. In the 2022 landmark decision Thaler v. Vidal, the Federal Circuit ruled that U.S. patent law requires inventors to be natural persons. This meant that DABUS, an AI system credited with creating novel inventions without human input, could not be listed as the inventor.
This decision reinforced the current legal consensus: only humans can be inventors, even if AI is the tool used. However, this precedent sidesteps the deeper implications AI raises for patent law.
Redefining Obviousness in the Age of AI
The KSR Standard and AI’s Expanding Role
In the 2007 decision KSR International Co. v. Teleflex Inc., the U.S. Supreme Court emphasized that patents should not be granted for advances that “would occur in the ordinary course” of development. The Court held that the fictional “person of ordinary skill in the art” (POSITA) is a person of “ordinary creativity,” not a robot.
But in today’s world, that fictional person likely does use AI tools — and those tools are anything but ordinary.
What Happens When AI Becomes a Common Tool?
AI’s use in innovation introduces a profound shift in the definition of what is obvious. AI can combine scientific disciplines, simulate millions of iterations, and suggest optimized solutions—tasks that would take teams of human experts months or years to perform.
Take the example of biotech companies using AI to discover new protein structures. AI can model, generate, and simulate thousands of protein variations in hours. What once would have been groundbreaking is now a routine, optimized output using AI tools.
This means:
- Inventions derived using widely available AI tools could increasingly be considered obvious.
- Courts and patent examiners may consider AI a standard tool that any skilled person is expected to use.
Case Study: AliveCor v. Apple
In a 2023 case, AliveCor sued Apple, alleging that Apple’s smartwatches infringed patents covering machine learning algorithms for detecting arrhythmias via ECG.
Apple countered by arguing that these patents were invalid due to obviousness, citing prior art that used ML to analyze ECG data, though not explicitly for arrhythmia detection.
The Federal Circuit sided with Apple. It ruled that combining machine learning with ECG data to detect arrhythmia was an ordinary creative step — something a person skilled in the art would naturally do. The court concluded that, even back in 2013, it would have been obvious to integrate ML.
This ruling sets a powerful precedent: If AI or ML is commonly used in the art, applying it—even in new ways—may not be patentable.
Pioneer Patents in the AI Era
Not all AI-driven inventions are destined for rejection. In fact, pioneer patents — those that represent foundational breakthroughs — could be strengthened by AI.
AI and the Enablement Requirement
The enablement standard, reaffirmed in the Supreme Court’s Amgen v. Sanofi decision in 2023, requires that patents teach others how to make and use the full scope of the invention without undue experimentation.
This standard has historically limited broad patent claims for pioneering inventions. For example, discovering a new class of proteins that treat disease is remarkable—but unless the patent teaches how to make most members of that class, it fails the enablement test.
AI could change that.
- Generative models can rapidly simulate all permutations of a new molecule class.
- Predictive models can identify variants likely to work, based on a few examples.
Thus, inventors can use AI to supplement their disclosures, enabling broader claims. This turns AI into an ally for patent enablement.
The Double-Edged Sword of AI in Patent Law
As AI becomes integral to research and development, its effects on patentability will be both limiting and empowering:
Limiting Factors:
- More inventions may be considered obvious because AI makes them easier to achieve.
- Patent claims based on mere AI application may struggle to meet the non-obviousness standard.
- Inventors must anticipate that AI is part of the hypothetical POSITA’s toolkit.
Empowering Factors:
- AI enables faster, broader experimentation, supporting enablement for complex inventions.
- Pioneering discoveries may be better documented and explained with AI assistance.
- Patent drafting quality can improve as AI aids with claim construction and data visualization.
The Future Legal Landscape: Evolving Standards
The legal community is now grappling with how to integrate AI into the patent framework. Several forward-looking trends include:
- AI-Aided Inventorship Guidelines – As AI involvement grows, new policies may clarify when human-AI collaboration constitutes true inventorship.
- Revised POSITA Standards – Courts and examiners may be required to assess what AI tools were available to a skilled person at the time of invention.
- AI as an Enabling Disclosure Tool – Patent filings might include AI-generated data to meet enablement.
- Regulatory AI Patent Databases – Patent offices may eventually maintain registries of accepted AI tools and datasets to guide prior art and enablement assessments.
International Perspectives on AI and Patentability
Globally, countries are taking different stances:
- European Patent Office (EPO): Like the U.S., EPO requires inventors to be natural persons but recognizes the use of AI tools.
- China: Recognizes AI as a research tool but has emphasized human oversight in inventions.
- Australia and South Africa: Have entertained AI-generated inventorship more flexibly but face legal uncertainty.
These differences could lead to forum shopping, where inventors choose jurisdictions based on their AI policies.
Real-World Implications for Innovators
Startups
Startups using AI tools must carefully document human input to preserve inventorship claims. They should focus on applying AI in novel ways rather than relying on AI-generated results alone.
Enterprises
Larger corporations must reassess their IP strategies, considering that AI-made discoveries might not be protectable through traditional patents. Emphasis may shift toward trade secrets or AI-assisted disclosure for broader patent scope.
Universities and Research Labs
Academic institutions should train researchers on the nuances of AI in patentability and establish new standards for inventorship documentation when AI is involved.
Conclusion: A New Patent Era Begins
AI is not just a disruptive tool—it is now part of the foundation of modern innovation. As such, it is naturally affecting patentability in fundamental ways.
- Obviousness will increasingly reflect AI’s capabilities.
- Enablement may be enhanced by AI’s ability to simulate and expand on initial discoveries.
- Inventorship definitions must adapt to accommodate human-AI collaboration.
The patent system must evolve to recognize AI not just as a tool, but as a force that is reshaping the nature of invention itself. Innovators, lawyers, and policymakers must work together to ensure that the next generation of creativity is not stifled—but rather, empowered—by the machines helping to fuel it.
As we move into this new era, the key question will no longer be whether AI is involved—but how its involvement shapes the novelty, non-obviousness, and enablement of what we create.