Patent practitioners may be overlooking one of the greatest value-adds of generative artificial intelligence tools for patent work – the ability of these tools to act as insurance.  

In the patent space, more and more practitioners are using generative artificial intelligence tools by the day.  The general consensus is that these AI tools offer two primary advantages: (1) speeding up patent work by automating repetitive tasks, and (2) preparing first drafts patent practitioners refine into a final product.  Both of these are real.  However, using these tools to have your back may be even more important, in three key ways.  

Avoid Missing Arguments

As most patent prosecutors, I pride myself on being able to compare my clients’ inventions to the prior art and identify distinguishing factors of the invention compared to the prior art.  After years of practice, I feel like I am highly skilled at figuring out which arguments will convince patent examiners that the claimed invention is patentable over the prior art.  In my view, crafting these arguments is an art form that I continue to improve upon year-by-year.  But I am imperfect.  Sometimes I miss arguments.  Occasionally, those arguments turn out to be the winning arguments.  If I am being completely honest, there have been times in my career when I was preparing a second response to an Office Action issued by the USPTO and came up with an argument that I should have included in the first response to Office Action.  I know I am not alone in this imperfection.  

One way that generative AI patent tools can act as insurance is by minimizing how often patent practitioners miss arguments that distinguish an invention from the prior art.  Currently available tools can analyze prior art in comparison to both the claims and the specification and identify differences between the invention and the cited references. Some of these differences are not helpful for advancing prosecution and some are blatantly clear to an experienced patent practitioner.  The most helpful insights tend to fall somewhere in between and can reveal angles that may otherwise be overlooked.

This benefit becomes especially clear when using step-by-step prompting.  Like most generative AI tools, the first output is rarely the best output.  Optimizing the helpfulness of an AI tool to distinguish the prior art can involve asking the tool to compare the claims to the prior art element-by-element.   If you have already analyzed the prior art and formulated your arguments, this element-by-element comparison should not take long and may not result in additional arguments.  But it can help ensure that you are not missing an argument.  

 It can also encourage deeper thinking about your current arguments, which can lead to tweaks that ensure you are putting forth the best work product for your clients.  Think of it like proofreading your work.  Sometimes when reading through your arguments one last time you catch important typos or clarify an argument.  Sometimes you do nt.  But overall, proofreading improves your work product.  Using AI patent tools to check for further arguments can provide a similar benefit.

Add Further Language to Patent Applications

It is not uncommon during patent prosecution to wish that a patent application contained few more sentences, or even a few more words, describing a feature that could have distinguished the invention from the prior art.  I have had this experience when prosecuting patent applications that I drafted myself, and when prosecuting patent applications that other patent attorneys drafted.  It is especially true when an invention is important and you are filing multiple continuation applications.  In later stage continuations, practitioners often try to thread the needle between infringeability and patentability/validity and a few extra words here and there can become extremely valuable.

Another way that generative AI patent tools can act as insurance is by making sure that features of an invention are described in greater detail.  AI tools can quickly generate additional language describing a feature.  If you write one sentence about a feature, it can be misinterpreted by the patent examiner, especially because patent examiners at the USPTO interpret language under the broadest reasonable construction rubric.  AI tools can expand that single sentence into several sentences, which can be extremely helpful in preventing the misinterpretation of a feature by the patent examiner.  It can help you argue that the examiner’s interpretation is unreasonable in light of the specification or provide additional options for amending the claims to overcome a prior art rejection.

For example, you could prompt an AI tool to expand a single sentence into ten sentences and then refine those sentences manually and/or through further prompting.  You can also ask the AI tool to describe the same concept in several different ways and pick and choose language that is helpful for adding to your patent application.

This approach can be especially helpful when writer’s block sets in or when drafting a patent application on technology that you have less experience in describing the underlying concepts.  

Training new patent practitioners

Being a patent attorney or patent agent is a difficult job, particularly for newcomers.  Training new patent practitioners can be time consuming and stressful because you want the work produced for your clients to be high quality, but you do not want to be charging your clients for double work. 

Another way that generative AI patent tools can also provide insurance is by making sure that novice patent practitioners are able to produce higher quality work more efficiently.  Of course, experienced patent attorneys should never relinquish their responsibility to review work directly and to sit down face-to-face with junior patent attorneys.  But AI tools can help new patent practitioners get more words in writing, which can be extremely difficult for newcomers.  Patent drafting and prosecution is a trial-and-error profession, but fear of making mistakes can be paralyzing.  AI patent tools can help reduce that barrier by assisting new practitioners in formulating arguments and describing inventions greater detail, allowing them to learn and improve more quickly.

Conclusion

In summary, beyond speed and drafting assistance, one of the greatest value of AI patent tools lies in their ability to improve the completeness and quality of patent work by acting as a safety net for patent practitioners.  These tools can reduce missed opportunities to strengthen patents during both drafting and prosecution. In that sense, generative AI can provide a valuable form of insurance for patent attorneys.