The advent of generative artificial intelligence has sent shock waves across industries, from the technical to the creative. AI systems that can generate viable computer code, write news stories and spin up professional-looking graphics have inspired countless headlines asking whether they will take away jobs in technology, journalism and design, among many other fields.

And these new ways of doing work and making things raise another question: In the era of AI, what does it mean to be an inventor?

Among technologists who build digital tools or programs, it is increasingly common to use AI as part of design and development processes. But as deep learning models flex their technical muscles more and more, even highly skilled researchers who are using AI in their work have begun to express concerns about becoming obsolete.

There is much debate about whether AI can augment human creativity, but emerging data suggests that the technology can boost research and development where creativity typically plays an important role. A recent study by MIT economics doctoral student Aidan Toner-Rodgers found that scientists using AI tools increased their patent filings by 39% and created 17% more prototypes than when they worked without such tools.

While this study indicates that AI seemed to help humans be more productive, it also showed there was a downside: 82% of the surveyed researchers felt less satisfied with their jobs since implementing AI in their workflows. “I couldn’t help feeling that much of my education is now worthless,” one researcher said.

This emerging dynamic leads to a related question: If a scientist uses AI in order to build something new, does the output still qualify as an invention? As a legal scholar who studies technology and intellectual property law, I see the growing power of AI shifting the legal landscape.

Natural persons

In 2020, the United States Patent and Trademark Office refused to list the AI system DABUS, which purportedly designed a food container and a flashing emergency beacon, as an inventor on patent applications. Subsequent court rulings clarified that under current U.S. law, only humans can be listed as inventors, but they left open the question of whether inventions developed by scientists with the help of AI qualify for patent protection.

The concept of inventorship and legal protections for inventions have deep roots in the U.S. The Constitution explicitly protects the “exclusive rights” of authors and inventors “to their respective writings and discoveries,” reflecting the framers’ strong conviction that the state should protect and encourage original ideas.

U.S. law today defines an inventor as a natural person who has conceived of a complete and operative invention that can be used without extensive research or experimentation. An inventor must do more than follow routine instructions – they must make an intellectual contribution in producing something novel.

That contribution can be a key idea that sparks the invention or a crucial insight that turns the concept into a working product. If a person’s input is routine or just explains what’s already known, they are not an inventor.

Role of AI

To what extent can or should AI become part of the invention process? The release of AI applications such as ChatGPT in 2022 introduced the public to large language models and sparked renewed debate about whether and how AI should be used in the inventive process. That same year, the U.S. Court of Appeals for the Federal Circuit heard a case that tested whether AI could be named as an inventor on a patent application.

The court concluded that under U.S. law, inventors must be human beings. The ruling reaffirmed the idea that Congress intended to encourage human beings, not machines, to invent. This idea remains foundational to current patent policy.

In light of the court’s decision, in 2024 the United States Patent and Trademark Office updated its guidance to clarify the role of AI in the inventive process. The guidance reaffirms that an inventor must be human. However, the Patent and Trademark Office explained that the policy did not preclude inventors from using AI tools to assist in the research and development of inventions. This approach acknowledges how the rapid development of AI technologies has allowed researchers to make exciting breakthroughs.

Policymakers seem to understand that if the U.S. is to continue to lead the world in innovation, the mythology of a sole inventor toiling away in a garage and relying on pure intellect must evolve to account for the value of AI tools that research has proven make humans more productive.

Nevertheless, since only human beings can be named as inventors on a patent, current policy does not quite answer the question of who or what should get credit for doing the work. Despite a growing trend where researchers are expected to disclose whether they’ve used AI tools, for example in academic papers, the U.S. patent system makes no such demand.

Regardless of AI’s role in the research and development process, a U.S. patent will list only the names of human inventors so long as those humans made a significant contribution to the invention. As a result, current policy is not concerned with how to recognize the contributions of AI. AI is considered a tool like a microscope or a Bunsen burner.

Personal ingenuity in the age of AI

Given this shifting legal landscape, I see that U.S. innovation policy is at a crossroads. The Patent and Trademark Office’s guidance reaffirming human inventorship and simultaneously embracing AI as an innovation tool is only a year old. It is unclear how the Trump administration’s forthcoming action plan to “enhance America’s global AI dominance” will affect this guidance.

Some observers expect the rate of scientific discovery to increase dramatically with the assistance of AI tools. But if the majority of those same productive researchers enjoy their jobs less, is the act of inventing being encouraged as the framers envisioned?

Current U.S. policy attempts to strike a balance and recognize the concept of personal ingenuity, stemming from the principle that for an invention to be patented in the U.S., a human must have led the way. Yet the guidance also implicitly acknowledges that AI can lend a helping hand in modern research and development. Whether and how policymakers maintain this balance – and how leaders in industry and science respond – will help shape the next chapter of American innovation.

W. Keith Robinson does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.