Artificial Intelligence (AI) is here to stay and will likely affect every aspect of our lives. When it comes to patents, there are a plethora of AI tools to assist with docketing, invention capture, patent application drafting, patent prosecution, valuation, invalidity analysis, and so on. More patent AI tools are being introduced almost daily.
Here’s my problem with the patent AI tools I’ve reviewed: They don’t help me learn. They take what I give them and make it look like a patent application or an office action response. So, if I don’t know what the result should be, and I rely on the accuracy and analysis of AI, I’m a disbarment waiting to happen.
It takes legal and technical skills, experience, and knowledge to write a good patent application and to appropriately prosecute it so that a quality patent issues (i.e., one that can be enforced). Increasing one’s knowledge, skills, and experience is never-ending. They have to grow with each new invention or the resulting patent application will suffer.
With respect to patents, knowledge is about knowing what to do and write and, more importantly, it’s about knowing what not to do or write. I have a saying, “there is no one right way to write patents, but there are thousands of ways to screw them up.”
I have yet to see any AI tools that address my concerns. What I have seen are AI tools that use LLMs which incorporate all, or almost all, issued and pending applications. There is no separating good from bad: There are no metrics for determining good patent practice, there are no metrics for identifying bad patent practice, there are no metrics to determine enforceable claim quality, and so on.
If an LLM includes the good, the bad, and the ugly, so will AI’s answers. If patent practitioners cannot tell “AI gold” from “AI garbage”, then all AI will do is help the practitioner write crappy patents faster. This helps no one and hurts the entire patent ecosystem.
For AI to help patent practitioners: 1. It needs to help practitioners learn more information faster, 2. It needs to compliment and increase human intelligence, not replace it, 3. It needs to leverage its data processing abilities to enhance human decision making, and 4. It needs to make sure that all critical decisions are made by humans, not AI.
**Please note that this article is not legal advice; it is not a legal opinion; nor should you rely on it as legal advice or as a legal opinion. This article merely expresses the author’s general thoughts on a topic regarding the business of patents. Nothing in this article establishes any form of an attorney-client relationship between you, the reader, and the author of this article.
