3L, Pepperdine University Rick J. Caruso School of Law
Synopsis:
I used Lexis+ with Protégé constantly. As a summer associate, my goal was to leave no stone unturned when it came to providing my findings, to give my attorneys a full picture of any topic I was briefing them on. Lexis Protege was my secret weapon in casting a wide net to achieve this objective, which was especially helpful on assignments where I wasn’t just focused on authoritative in-jurisdiction case law but also looking far and wide for relevant persuasive state and federal case law as well. On one project, Lexis Protege helped me uncover a line of federal cases that meant the difference between only recovering a nominal amount for our client, versus a colorable argument for getting disgorgement in a giant contract, the monetary difference being millions back on the table. Lexis Protege is a resource everyone must take advantage of.
3L, SUNY at Buffalo
Synopsis:
AI was a significant part of my workflow this summer, but primarily as a tool for accelerating work rather than replacing legal judgment. I used AI most heavily at the beginning and end of assignments. At the beginning, it helped me organize unfamiliar issues, identify potential statutes, regulations, cases, and search terms, and turn a complicated assignment into a workable research plan. I also often used AI to help organize the authorities and facts into a first draft of a memorandum or other work product. I also found AI especially useful during revision. I used it to identify redundancy, tighten legal writing, compare a draft against the governing statute or case, check whether individual propositions were actually supported by the cited authority, and flag factual or citation issues that warranted another look. For document-heavy assignments, it could help me compare versions, summarize lengthy files, extract relevant information, and spot inconsistencies that would otherwise take much longer to find manually. AI is much more useful when treated as something that must be carefully supervised rather than ever really trusted. In legal work, I still had to read the cases, statutes, regulations, and underlying documents myself and verify citations and factual statements before relying on them. AI could produce a polished answer that was subtly wrong, so verification was essential. Used that way, however, it materially increased both my speed and the amount of attention I could devote to the substantive legal analysis. Instead of spending as much time on mechanical tasks such as reorganizing notes, comparing drafts, or cleaning up a first pass, I could spend more time checking the record, reading the authorities, and refining the final work product. By the end of the summer, I was using AI more as an interactive research and drafting tool.
3L, Southwestern Law School
Synopsis:
Our firm actively experimented with AI to draft responses to cases. We also used AI to do legal research for a wide variety of legal areas (ex: ethics, entertainment, landlord/tenant). We also compared several AI products against each other to see which one gave the best response. Anything drafted by AI was reviewed and edited by us students, law clerks, and the attorney.
a
3L, George Washington University School of Law
Synopsis:
I used AI particularly in discovery requests including but not limited to requests for production of documents, request for admissions, and interrogatories. AI was most helpful in those areas because it covered very specific discovery instructions that are difficult to remember each time you draft another discovery document. AI was also helpful in searching a wide variety of case law on a number of niche issues that come up very rarely / are not touched on by the jurisdiction at all. AI does a good job of suggesting where you can look for some information and how to synthesize cases / what to suggest to the court as how to interpret a law.
3L, Vermont Law School
Synopsis:
This summer I interned at the Miami Public Defender's Office, where I drafted motions in limine and research memoranda for my supervising attorney. AI was not an optional add-on there. The office actively promoted AI use, and attorneys worked with usage expectations tied to AI, so I learned the tools the way I learned the docket, by using them on real work with real deadlines. My workflow settled into three distinct stages, each with a different tool doing a different job. I used AI for front-end research, identifying the governing standard on an evidentiary issue, pulling the case law, and orienting myself in an area quickly enough to start writing the same day rather than the next one. I used Claude for drafting assistance and, more importantly, for citation discipline, running every case I intended to cite back through the actual opinion and holding to confirm that the proposition in my brief was the proposition the court actually stated. I used Lexis+ with Protégé at the back end, analyzing the finished document, checking quotations against the source, and Shepardizing before anything went to my supervisor. The lesson that mattered most was that speed and verification are separate problems, and the second one does not solve itself. AI compressed the research and first-draft stage substantially. It did not reduce the time I spent confirming that a case said what I claimed it said, and it should not. Hallucinated citations are a real risk, and the only reliable defense is reading the opinion. Treating verification as a mandatory step rather than a discretionary one is what made the speed usable. What I take into my final year is that these tools change where the work sits, not how much judgment it requires. The output still carries my name and my supervisor's, and the responsibility for every proposition in it stays where it has always been.
3L, University of South Carolina
Synopsis:
I use AI largely as a research starting point. It's better at searching databases or the internet and gives me good background. I also use it to look up good search terms for AI. Also drafted one litigation doc, but it was overblown and I edited it a good bit.
3L, University of Akron
Synopsis:
This summer, AI has been a great help at both of my jobs. My boss is writing a research paper about AI as a publisher and I uploaded the cases to AI to find the best citations and stay organized. I also used it at my fellowship to assist on agency comments and help get summaries of useful information from previously filed comments. It cuts out a lot of time as a researched and bluebooker.
3L, University of Houston
Synopsis:
I primarily use AI to conduct preliminary research. Typically, I will describe the legal issue(s), ask what my guiding research questions should be, and ask AI to help me find sources, like relevant case law and statutes. The AI will provide me with a great starting point, and then I verify the information that it provides me by double-checking the statutes and cases on trusted websites. I then move on to creating the actual work product. In the revising and editing stage, I will upload the work product (or portions therof, if confidentiality is a concern) and ask the AI to check it for me. I'll generally ask it to check for specific things, like fact-checking, Bluebook citation formatting, verifying case law and statutes, checking the soundness and persuasiveness of my legal arguments, etc., and then I will go back and make revisions based on the advice AI provides. By using AI to gather preliminary research and polish my final product, I've significantly improved the quality of my legal work.
3L, University of Houston
Synopsis:
I elected not to use AI very much, if at all. I wanted to try to hone my own legal analysis and research skills. But, the other intern I worked with it used it almost primarily for every assignment we did. One day, I was working on an assignment for our supervising attorney. The attorney came in and asked me a question about the assignment, and the other intern said "I bet AI can find those answers for you." So, she inserted the legal question into AI. It was a battle of the minds (if you count artificial intelligence as a mind). My fellow anti-AI law students will be glad to know that while the AI is pretty good, I found all the cases it found already (and I didn't make up any fake case quotes!)
3L, Texas Tech Law
Synopsis:
AI became one of the most useful tools I learned to work with during my summer position at a litigation firm. I primarily used it as a starting point and organizational tool rather than a replacement for my own legal analysis. For example, I used AI to help organize deposition transcripts and appraisal reports, identify issues to investigate further, develop research questions, summarize large amounts of discovery, and create initial outlines for motions and other assignments. It was especially helpful when I was given an unfamiliar issue and needed to quickly understand the terminology, relevant legal concepts, and questions I should be researching. The biggest lesson I learned was that AI works best when paired with traditional legal research and independent verification. I still reviewed the underlying documents, checked cases and authorities myself, and made the final legal judgments. By the end of the summer, I viewed AI less as something that “does the assignment” and more as a tool that helps me get oriented faster, spot issues I might otherwise miss, and spend more of my time on the substantive legal analysis. For a law student working on real litigation matters for the first time, that made a significant difference in both my efficiency and the quality of my work.
3L, University of Florida
Synopsis:
I did not use A.I. at all for my internship this summer. Everything that we were helping petitioners or respondents with would go directly to the court, most of the time that same day, so it was critical that we checked that all the fields were filled out correctly on their forms. The most automation we used was to populate information fields (e.g., petitioner name, respondent name, address, case number). All declarations were drafted by interns based on an interview with the self-represented litigant, and reviewed for grammar and formatting by staff attorneys.
3L, University of Kansas
Synopsis:
This summer I used AI to help draft contracts and improve language and wording for contracts, and complaints/answers. The AI was helpful to help me format and expand on ideas. It was helpful to know what areas needed expanding and for contract drafting it was helpful to draft language I had never seen before.
3L, Southern Methodist University
Synopsis:
When I was done drafting something for a partner, I used AI to "grade" my work. I instructed it that they are a partner and I am a summer associate, and to identify any holes in my analysis, as well as basic grammatical errors and overall structure. It definitely helped me whip my papers into shape for submission.
a
3L, University of San Francisco
Synopsis:
This summer, I was the only law clerk at my firm and was starting out in family law, a highly specialized area that can be difficult to navigate as a beginner. I often used AI as a starting point when I encountered an unfamiliar issue, asking it to help me break down complicated legal issues, identify questions to research or organize my initial thoughts before diving into the case law. AI helped me get through the steepest part of the learning curve more efficiently, but it never replaced my legal research or judgment. I verified its responses against statutes and cases and made my own decisions about how the law applied. AI gave me more time and space to develop legal research.
3L, Santa Clara University
Synopsis:
This summer one of my assignments was to write an opposition to a motion to compel arbitration in a complex construction case. I began my research and was doing really well until I hit a wall. I spoke to the partner I was working on the assignment with and he suggested I try to prompt Lexis AI Protege and get case law to support my argument. I did get case law and I then was able to go to the case opinion and read to make sure it was in fact correct and made my argument stronger instead of weaker. AI helped me when I hit a wall during my research phase for this specific complex opposition to a motion to compel arbitration.
a
3L, Northwestern University
Synopsis:
I used AI to conduct legal research because it helped me streamline my searches. In particular, I used an AI deep research tool to search for a compilation of employment law statutes. I found it helpful to use this AI search function because it served as a great starting point for collecting information. It also allowed me to find gaps in my thinking by pulling up more statutes than I had originally considered.
a