P&L executives concerned with managing legal risk and controlling legal costs should know that artificial intelligence (AI) promises greater accuracy and lower costs in litigation tasks. If and when the legal industry adopts the technology.
And last month some of this promise appears to have been realized in some concrete, and economically accessible, terms. Again, conditional on actually acceptance and use of this AI.
For a business, litigation is (usually) a huge waste of money. And a large chunk of this wasted money goes to formal requests that lawyers make to judges. Requests that they and their adversaries spend lots of time (read “money”) fighting over in front of a judge.
It’s called “motion practice”: Your honor, please dismiss this case; please exclude this testimony; please make my adversary give me the papers in their files that I want to look at; etc.
In a legal industry where the number of hours billed (usually) defines value, this typically results in each law firm assigning not just a litigation veteran whom they put in charge of the case — but also multiple, less-experienced attorneys to maximize those hours billed.
How do these less-experienced attorneys fit in here? Happily (for the legal industry’s business model at least), “motion practice” requires lots of what those recent law grads were trained to do when still in law school: Researching cases, statutes, and rules; and creating “briefs” that describe why those legal authorities require that their client’s requests (via this “motion practice”) should be granted.
Late last month, Casetext, a legal tech firm known for its artificial intelligence (AI) research tool “Case Analysis Research Assistant” (CARA), announced “Compose”, their automated brief-writing product.
In other words, this legal tech firm had already created “CARA” to automate legal research. Ross Intelligence and Westlaw Edge as well offer competing automated legal research products.
“Compose” takes this AI-based functionality another step by automating the writing of “briefs” for “motion practice”. It appears to be the first such system commercially available to the legal industry.
As described by my favorite legal tech commentator (Bob Ambrogi):
” … To use Compose, a lawyer begins by entering basic information about the brief, such as the nature of the brief, the jurisdiction, and the lawyer’s position for or against the motion.
“Compose then presents the lawyer with a treatise-like outline of what Casetext says are all the available arguments for that motion, as well as the legal standards or rules applicable to each argument.”
Can the AI system “write the brief”? Is it a “robot” in that literal sense? No.
But this (and, soon, additional such systems, I expect) provides the first drafts that an experienced lawyer can then tweak and perfect. (And my initial impression is that “Compose” will be financially accessible to smaller law firms as well as to the larger ones.)
If initial reports are borne out, “Compose” will enable a materially lower level of grunt work for which law firms currently bill their clients large fees.
Therefore, a materially lower level of such fees. If and when it (and similar future systems) are adopted by the legal industry.
Law firms won’t make good use of this AI system unless they’re willing to displace the hours billed by their young lawyers in these “grunt” level tasks. I’m not holding my breath for immediate, widespread adoption by law firms.
Such adoption by law firms, if and when it happens, is likely to require a push from the client companies that hire those law firms.
Realizing AI’s promise of greater accuracy and lower costs in this area calls for more than just AI savvy. It will require the forceful use of client corporations’ purchasing power with their law firms.