Technology’s Effect on Advocacy
accelerate reasoning. It cannot assume professional responsibility.
imposed discipline. You could not casually claim that a case stood for a proposition without locating the opinion, reading it, and checking what happened to it afterward. As a law student, I became addicted to Boolean searching. I loved the precision of it. One additional connector, proximity term, or factual phrase could uncover a case that a broader search had missed. Westlaw and Lexis transformed the research landscape. They made it possible to locate authorities faster, identify negative treatment more efficiently, and explore lines of precedent that once would have required hours of manual research. Yet Westlaw and Lexis never relieved lawyers of the responsibility to read the cases. A headnote could point me in the right direction, but it could not tell me whether the language came from the holding, dicta, a concurrence, or a party’s argument. A red or yellow flag could warn me that a decision had been criticized or limited, but it could not determine whether that treatment affected the proposition for which I intended to cite it. The platform assisted my research. It did not exercise my judgment. Artificial intelligence is the next stage of that evolution. It can search, summarize, compare, organize, and generate with breathtaking speed. It can identify authorities addressing an emerging issue, locate recurring themes in hundreds of pages of testimony, or compare multiple opinions in seconds. Those capabilities are enormously valuable to plaintiff’s lawyers, particularly those practicing against institutions and defense firms with far greater resources.
How I Actually Use AI to Brief Motions
I do sometimes ask Westlaw’s AI tools or another research platform to suggest cases on a developing issue. That is an increasingly useful part of modern research. A well-framed question may surface an authority I would not have found through my first Boolean search, especially where the terminology is evolving or different courts describe the same issue in different ways.
But a suggested case is not a verified case, and an AI summary is not legal analysis.
Once a potentially relevant authority appears, I pull the opinion. I read it. I examine the facts, the procedural posture, the standard of review, and the precise language of the holding. I check whether later cases have limited, distinguished, or questioned it. Only then do I decide whether it belongs in my argument. After I identify the cases that matter, I often feed the actual opinions back into AI. That is where the technology becomes especially useful. I may ask it to compare several cases, identify where their reasoning converges, isolate meaningful factual distinctions, or explain whether my proposed rule is broader than the authorities support. I may ask which case opposing counsel is most likely to rely upon, whether the cases reveal a split in analytical approach, or whether a recent decision changes how an older case should be understood.
I do not merely ask AI to confirm that I am right. I ask it to tell me why I may be wrong. Some of my
But AI is not the next generation of legal judgment. It is the next generation of legal tools. Technology can
17
Consumer Attorneys of California
FORUM September/October 2026
Made with FlippingBook - Online catalogs