Prof. Jeffrey Lefstin, Vanguard Award winner in IP, addresses patents in an era of transformation 

This is a headshot of Professor Jeffrey Lefstin, wearing a suit and glasses in front of a white background.

Professor Jeffrey Lefstin is the winner of the 2026 Vanguard Award from the California Lawyers Association for his work in intellectual property. Trained as a biochemist, his scholarship includes the history of intellectual property law and its application to new modes of scientific discovery and invention.

 

  • Professor Jeffrey Lefstin won the 2026 Vanguard Award from the California Lawyers Association for his work in intellectual property. 
  • Lefstin earned a Ph.D. in biochemistry, and uses that training in his teaching and scholarship.
  • Lefstin’s scholarship includes the history of intellectual property law and its application to new modes of scientific discovery and invention.

  

Professor Jeffrey Lefstin won the 2026 Vanguard Award from the California Lawyers Association for his work on scholarship and public policy related to intellectual property. In a Q&A, Lefstin, who earned a Ph.D. in biochemistry before turning to the law and who has testified before Congress about patent law, comments on his work, challenges arising from artificial intelligence and applying IP to rapidly advancing technology.  

Q: How does your background in biochemistry inform your work in IP? 

Supreme Court Justice Joseph Story once called patent law “the metaphysics of the law.” Inventors create specific things. But patent law has to give inventors right beyond the exact things they create – otherwise the patent would be worthless. The hardest problems in patent law are how to establish the category of possible things that the inventor has rights over, and how to define those categories in words. A biologist – particularly one with an evolutionary bent – naturally thinks in terms of hierarchical taxonomies.  

The very first day in my patent law class in law school, our professor used the example of a cloned insulin gene to illustrate the scope of patent claims. I immediately recognized that the hard question was how far the inventor’s rights would extend: if you’ve cloned the gene from a rat, can you assert rights over the mouse gene? Other mammalian genes? The human gene? I’m still working on those problems. 

Q: What are the most pressing issues in IP? 

Issues with generative AI are at the forefront of IP right now. Courts around the world have readily concluded that an AI can’t be the inventor of a patent, or the author of a copyright, in its own right. The hard questions that haven’t been yet resolved are to what extent a human aided by AI can claim to be an inventor or an author. How much human creativity is required? Can you claim IP rights in the AI’s output, or only on the human contribution itself? Must an author or inventor disclose the use of AI to a patent or copyright office? We’ve never required complete human control – after all, a digital camera makes some of the decisions for you when you press the shutter button. But we’ve not doubted that the photographer is an author entitled to copyright. Generative AI requires us to question how far that reasoning extends.  

Another big question is whether using literary works to train AI models is permissible under copyright law. Some of the early court decisions have reasoned that because humans can read books and absorb their content without further permission from the copyright holder, AIs ought to have the same privilege. But that’s a very formalistic kind of reasoning. It was a deliberate design choice of our legal system to allow humans to read and ingest content, and to build on the author’s ideas, without permission from the original author. Do we want the same for AI? 

Less talked about are some of the IP risks involved in generative AI. For example, if you use a large language model (LLM) to help write your code, you may be incorporating other people’s IP into your work without even realizing it. 

In patent law, the scope of what is patent-eligible subject matter has been dramatically reshaped by Supreme Court decisions over the last decade. Patent-eligible subject matter is about the basic boundaries of the patent system: for what kind of inventions might we even consider granting a patent? The drafters of the 1952 Patent Act intended the doctrine of patent-eligible subject matter to be a very coarse filter, with the real questions of patentability to be decided by more specific provisions of the Act. But the Supreme Court, and the lower courts in the wake of the Court’s decisions, have excluded from the patent system essentially allinventions based entirely on new scientific discoveries, as well as most inventions based on advances in computational techniques. The result is that, after years of pushing the rest of the world to increase standards of patent protection, the United States protects fewer areas of new technology than almost any other industrialized nation. These are the kinds of large-scale policy decisions that ought to be made by Congress. But the courts have backed us into this regime without reference to either the statutory text or Congressional intent. 

Q: What can your research tell us about today’s challenges in IP?   

One of my major research interests is the history of patent law. Many of the questions the courts are dealing with in cases on biotechnology or artificial intelligence – such as the proper boundary between unpatentable discovery and patentable invention – are the same ones the courts were addressing in the late 19th century, which was the formative period of American patent law. 

What’s striking to me is how reliant today’s Supreme Court has become on history to justify its decisions. In the patent-eligible subject matter cases I just talked about, the justices tell us they’re just applying the same principles the Supreme Court applied in its decisions on Morse’s telegraph or Alexander Graham Bell’s telephone more than 150 years ago, or the same principles the English Court of Exchequer applied in decisions on the hot-blast iron smelting process in 1841. 

But the Court has gotten the history 100% wrong. My own work has shown that those 19th-century patent cases stand for the completely opposite propositions than the Court has claimed. Back then, the Supreme Court was deeply engaged with both the technology and the law in patent cases. Today’s Court is much less engaged on both fronts. So the Court may imagine that it’s restoring patent law to its historical roots, but in reality it’staking the patent system in an entirely new direction, based only on its own economic hunches.