How AI is Laying Down the Law: Future
- Madison Shanfeld
- a few seconds ago
- 3 min read
Welcome to part-two of How AI is Laying Down the Law! In this article we will be focusing on the forefront of development for Legal AI with expert Nate Jedinak, Senior Director of Software, Data, and Innovation at Vorys.
His Thesis
In the modern day, legal AI is generalized to the practice of law, not to that of a single practitioner. When Nate seeks to address the question of how to make legal AI as efficient and productive as possible, he questions the vast generalization of AI.
He notes that when lawyers were asked what made the best patent lawyer or litigator, attorneys would often respond with a name, not a generic ability. Therefore, to make an AI as efficient and as useful as possible, one possible avenue is to model the AI after an actual person
Initial Work Towards Deeply Personal AI
Through a collaborative partnership with Stanford Liftlab, an innovative legal AI lab at
Stanford Law School, Nate is developing various advanced solutions to test, learn from, and practically apply the concept of personas. Initial work focuses on building deeply personal artificial intelligence modeled after the approach, knowledge, skills, and experience of a specific individual. In theory, these AI personas would replicate the knowledge, experience, and writing style of who they are built to represent.
Tackling Young Lawyers
One avenue of research is how personas could be used to enhance the training, development, and mentoring of early stage lawyers. Several systems being developed by Stanford, Nate, and his team may be incredibly helpful to early stage associates. Typically, associates will have a partner or a set of mentor partners who assign them tasks, like research, and who are available to answer a plethora of questions. However, partners become incredibly busy with their client caseload and responsibilities within the firm. Thus, they may not have enough time to answer all associate questions in verbose detail. Additionally, pedagogical skill varies among practitioners.
This assignment of difficult research played a part in how early stage associates learned in past decades. Now that even general-use AI has become more advanced, artificial
intelligence can often complete a majority of this workload. With AI completing this research, associates no longer learn as extensively through the research memo process. However, it is surmised that by applying the concept of personas, associates would be able to collaboratively research alongside their partners’ AI persona; asking questions, probing concepts and theories, and earning a verbose and instructive response without taking time away from the partners’ responsibilities. This allows associates to ask more questions, saving the most difficult concepts for human collaborative work, potentially creating opportunities for both partners and associates to become more efficient.
Beyond the Associates, Bionic Attorneys
This new concept in Legal AI is not only for aiding new lawyers; it is also incredibly beneficial for skilled practitioners. Nate aims to create technology that allows attorneys to feel “bionic”—like the best versions of themselves, removing distractions to allow more time for the most creative and impactful legal work.
One interesting aspect of research is consistency. Nate attests that technology can always review everything at 100%. Unlike human beings, technology does not care how many tasks are stacked up in one’s backlog, experience stress, will not catch the flu, become frustrated, lose steam due to hunger or dehydration, or allow other external factors to impact the cycle time devoted to a task. One notable early success has been partners using a persona-based redlining tool to review their own documents with their persona as a final quality check. Even if the document is hundreds of pages long, the AI persona can apply the same evaluative rigor from the first paragraph as the last.
Another avenue of early success and continued research is using a persona to bounce ideas off trusted coworkers or partners with different skillsets, without pulling others away
from important tasks or heavy caseloads. This can be helpful as attorneys cannot be experts in every arena of law. Utilizing AI personas may create opportunities for lawyers to get expert feedback on areas of law they may not specialize in, allowing them to generate the best possible outcome for clients. Nate explained this with the example of patent law. Patents are almost like telling a story, but in a foreign legal language. Patent attorneys tend to be extraordinarily detail-oriented even in the face of extremely difficult and disparate concepts. He says, “If you can apply the concept of a patent attorney persona to a non-patent document, you may come out with a better constructed, more persuasive document.” Another interesting possibility is applying a litigator’s persona to a contract: early research showed a specific litigator’s persona could point out risks in novel ways.
Running on a Handshake
According to Nate, attorneys are “relationship people,” and he sees that as a key skill well into the future. Nate moved to San Francisco in 2025 to immerse in the Silicon Valley atmosphere; he noted that in the Valley, an incredibly tech-forward environment, “everything still runs on a handshake.” Nate believes AI that supercharges what a lawyer can accomplish can allow lawyers to focus on and add unique personal value in the relationship aspects of their profession.