As a third-year law student, I’ve had a front-row seat to the artificial intelligence revolution in the legal field. This experience has helped me integrate AI into my research, writing and studying.

The rise of AI in legal education

With tools like ChatGPT, Lexis+ AI and locally hosted large language models, law students have ample opportunities to embrace this revolution, rather than resist it.

I remember when OpenAI took the world by storm, releasing ChatGPT in November 2022. In my first round of on-campus interviews taking place in January, I asked one of the interviewing associates if their firm (an AmLaw 100 firm no less) had any plans to incorporate AI into their workflow in the upcoming years. His response was, “Anyone who tells you they plan to use the technology is lying.” This was surprising to me, as I had already embraced AI, using it as a study aid for my December 2022 finals.

Indeed, AI quickly made headlines in the legal world, notably when a New York attorney was sanctioned in 2023 for failing to verify the accuracy—or even existence—of cases cited by ChatGPT. Next, LexisNexis launched Lexis+ AI in October 2023. Before that, I was using a lesser-known tool called Paxton AI.

AI’s role in legal research and writing

My interest in AI even led to a position where I was tasked with grading the accuracy of responses to legal questions. This role in particular served me well as I was studying because I would task the AI agent with questions about the material I was learning in my classes, offering corrections and feedback where it was needed.

One of AI’s greatest advantages is its ability to enhance the early stages of legal research. These tools can rapidly analyze and summarize vast amounts of data and, depending on the tool, provide citations to legal sources or newspapers that I can use as a springboard for further research.

This capability has streamlined my research process, enabling me to focus more on analysis and application, rather than finding the rule that applies to a specific case. For example, when I was working with the Tennessee attorney general’s Criminal Appeals Division, my first step in the research process was to place the general issue into Lexis+ AI to find relevant caselaw and winning arguments to incorporate into the appellate briefs I was drafting.

AI has also been an invaluable learning aid. Uploading PDF versions of assigned readings into AI models and asking for key takeaways and analytical checklists has helped me focus on the most relevant legal issues.

For example, in my class on civil rights actions, I upload law review articles into various models to highlight the unique contributions of the author to the discussion. This approach is particularly beneficial when dealing with intricate doctrines or unfamiliar areas of law.

Ethical and practical challenges of AI in law

As AI continues to develop, I am closely following its impact on legal writing. Some tools now allow firms to upload previous briefs, enabling AI to replicate the writing and citation styles of a particular office or practice area. This efficiency allows me to adopt the writing techniques used by top litigation attorneys in successful cases into my legal writing.

Looking forward to the future, I wrote another piece on the use of AI in determining the “plain meaning” of a statute at the time it was written. As judges trend toward the use of “original public meaning” to determine the meaning of statutes and constitutional provisions, using these tailored tools could be the future battlefield in resolving disputes.

Tools like Alphabet’s NotebookLM, which allows for the results to be based on specific uploaded sources, will allow attorneys to base their definitions on dictionaries, encyclopedia entries and newspaper articles from the time the statute was enacted, enable litigators to argue for one definition over another.

Balancing AI and traditional legal skills

That being said, being an early adopter of the technology requires a particular knowledge of how these systems work. For example, in my legal writing class during my 1L year—before the New York attorney was caught and sanctioned for using AI—I had the exact experience he did. But knowing about the risk of so-called “hallucinations,” I looked up and discovered the cited cases were indeed “made up” more often than not.

Indeed, even when using tools like Lexis+ AI, the citations provided often were not actually relevant to the question I presented the agent with. This experience underscores the importance of using AI as a starting point, rather than a tool for production as the technology is currently implemented.

Another issue I’ve encountered with the more general-purpose tools such as ChatGPT is the lack of citations. These tools often lead to challenges in verifying sources and ensuring the accuracy of the output. This limitation, when properly accounted for, has often increased the time I would otherwise have spent studying, particularly when asking for the answer to specific legal questions.

Confidentiality concerns also arise when using these tools for work product. The University of Tennessee recently updated its AI policy, and with a focus on clinic work, it noted “Users should assume that all consumer generative AI products make data publicly available unless otherwise indicated per explicit official agreement with the University of Tennessee.” This requires me to tone down the specific questions, making it difficult to find answers relevant to my research questions.

But as I found in researching my article for my information and data privacy seminar, the use of downloadable, locally hosted models prevents this data from being sent to any outside servers, ameliorating such concerns. As computer technology increases and as firms invest in local servers to run these models, these issues become less prevalent.

The future of AI in law and legal interpretation

I frequently ask myself if I’m overrelying on AI and hindering my development of critical legal skills. While AI does indeed provide quick answers to complex questions, it’s essential to engage with the material, rather than accept it at face value.

Balancing AI assistance with my traditional study forms has been crucial not only to ensuring a comprehensive understanding of the law but also to ensure that I am not relying on shortcuts that will be detrimental to my future practice.

It’s clear that AI offers valuable tools to enhance legal research and study. But it’s imperative to approach these tools with a deep understanding of how they work, along with strong human judgment to prevent their misuse and avoid confidentiality issues.

By striking this balance, I and other law students can harness the benefits of AI while mitigating its drawbacks, enriching our legal education, and preparing us for the ever-changing landscape of the legal profession.

Austin Gergen is a 3L at the University of Tennessee, and he is set to graduate in May with his JD and a concentration in advocacy and dispute resolution. After graduation, he will clerk for Judge Jeffrey Usman on the Tennessee Court of Appeals.

ABAJournal.com is accepting queries for original, thoughtful, nonpromotional articles and commentary by unpaid contributors to run in the Your Voice section. Details and submission guidelines are posted at “Your Submissions, Your Voice.”

This column reflects the opinions of the author and not necessarily the views of the ABA Journal—or the American Bar Association.