By Michael Langan1
Most lawyers talk about generative artificial intelligence (“gen-AI”) in terms of efficiency. How many hours can it save? But is that the right question?
Assume it traditionally takes you 40 hours to prepare a brief in support of a summary judgment motion. Now assume that, with gen-AI, it takes only 20 hours to produce that same brief (even after verifying every output).
Congratulations. You’ve doubled your productivity. But you’ve also kind of missed the point. The brief isn’t better. It’s simply the same brief, prepared faster.
The real question is, “What should I do with the 20 hours I just saved?”
A businessperson might snap, “Start the next case.” A lawyer would reason, “If the purpose of the brief is to be more persuasive than that of its competitor, then shouldn’t I spend the time making the brief more persuasive than usual?” After all, the purpose of a legal brief isn’t to finish it. It’s to win the motion.
The challenge, of course, is defining “better.” You may shrug and say that’s subjective. But is it really? In 2023, I reviewed the briefing of 58 summary judgment motions filed in the Northern District of New York during 2018. Approximately two-thirds of the time, the table of contents (“TOC”) of the winning brief contained more of seven objective elements than that of the losing brief.2
That finding is hardly surprising. (Wow, the better-written brief won.) But it points to an intriguing possibility: winning briefs may share objectively identifiable characteristics. If so, perhaps our conversations should be less about how to use gen-AI to save time and more about how to use gen-AI to improve quality through more intelligent revision. One way to do this is to use gen-AI to help revise our table of contents so that it possesses all seven objective elements that experience suggests correlate with success.3
Lawyers who use gen-AI only to save time may become more efficient. But lawyers who use gen-AI to improve the quality of their briefs may win more motions. (Unless Luddites up their game by improving their briefs the old-fashioned way.)
1 Mike is the career law clerk to United States District Judge Glenn T. Suddaby. The views expressed in this article do not necessarily reflect the views of Judge Suddaby or any judge in the Northern District of New York.
2 The seven elements were these: (1) which argument section of the TOC had more words (not including any subpoints setting forth the misplaced statement of the legal standard or relevant background); (2) which argument section of the TOC used more complete sentences versus fragments (other than in fifth level of outline specificity); (3) which argument section of the TOC used more points and subheadings in total (i.e., headings labeled “A,” “B,” “1,” “2,” “a,” “b,” and “i,” “ii”); (4) which argument section of the TOC more often used the words “claim,” “cause of action,” “count,” or “element”; (5) which argument section of the TOC expressed more causation (e.g., “because,” “as a result,” “consequently,” “resulting[ly],” “since” “as,” “given that,” “therefore,” “thus,” “in view of,” or “as such”); (6) which argument section of the TOC expressly refuted more concessions (e.g., “even if,” “assuming,” “although,” “even though,” “while,” or “despite”); and (7) which argument section of the TOC more often argued in the alternative (e.g., “in any event,” “in the alternative,” or “as a threshold matter”).
3 Of course, this correlation makes intuitive sense: when we better craft our table of contents, we improve the nature and order of all of the headings and subheadings throughout our brief, which in turn reminds us to keep the material in those sections restricted to the designated subject. This makes our arguments easier to understand, better supported, and more cogent.