Technology that improves the quality and efficiency of legal work will be stymied as long as hourly billing prevails among attorneys. At least that’s my belief.
Why do I say that?
- Hourly quotas for attorneys encourage more lawyer time per task;
- This motivates a proliferation of lawyers on any given task — each lawyer with their own hourly quota; and
- This proliferation leads to insertion of recent law graduates alongside fully qualified attorneys to do the routine & repetitive work for which law clerks or paralegals are suited — but billed to clients at several hundred dollars per hour.
A profession that gets paid more when its work takes longer, that gets paid even better when it assigns more people to do that work, and that charges clients for training its junior personnel — will lose money using a system that improves the quality or efficiency of that work.
Pretty straightforward, I think.
Last week I heard from a couple of others on this.
On December 6 I attended “Artificial Intelligence in the Enterprise (Legal Services Version)”, sponsored by the Law and Technology Initiative of Northwestern Pritzker School of Law and Northwestern McCormick School of Engineering.
Dr. Kristian Hammond is the Bill and Cathy Osborn Professor of Computer Science at McCormick, and the lead AI teacher and researcher within Northwestern University. He’s an expert on the application of AI to the professions — including law, journalism, business management, and medicine.
Kris gave a wide-ranging survey of AI’s fundamentals, with an emphasis on how its systems are applied to perform tasks that usually call for human intelligence — specifically, tasks now performed by attorneys.
To oversimplify, AI already offers greater accuracy and huge time-savings in a wide range of basic lawyer tasks. Enormous advances in the quality and speed of these tasks is within our grasp. If, that is, the legal profession actually adopts these systems.
At the end of a day of lucid (even as heard by a lawyer and business guy) lectures about AI, and how it can apply to law, Kris asked the class for our reactions.
I raised my hand, and said that I’d originally been skeptical that the legal profession’s business model posed a tangible impediment to tech adoption. But that conversations with founders of legal tech companies had convinced me that law firms’ dependence on time billed largely precluded substantial tech adoption. Time and again these founders (continue to) tell me that law firms reject AI (and other tech) systems’ increased quality and cost efficiencies because, those law firms’ representatives say, such systems would materially cut into their top line.
An attendee from a large law firm suggested a helpful modification of what I’d said:
“I find that lawyers’ resistance to tech adoption — especially as considered in light of the lawyers’ business model — varies according to the ‘margin’ inherent in the problem.” (Since most law firms are top line — and not net income — focused, he might better have said: “Revenue from hours billed”.)
Legal tech that can solve what he called a “high margin” problem — tasks that are a law firm’s bread and butter — “get no traction”. But legal tech that can solve a “low margin” problem — pain-in-the-neck tasks that must be done, but that they can’t charge much for — those “get traction”.*
After 20 minutes of discussing how the legal business model affects AI and other tech adoption in law, Kris said to the class:
“If it’s the case that the law firm business model is getting in the way, there is going to be something horrible that’s going to happen …. Essentially what happens is if there’s a way to do something faster, better, cheaper and with fewer hours, and firms aren’t doing that, then you’ll get spin-offs into that technology and competition from them.”
Roy Strom, a prominent reporter for Bloomberg Law who was present in the classroom, focused on Kris’ words about “something horrible” happening to law firms (“I was struck with a sense of foreboding in [Kris] Hammond’s warning”).
In an article entitled, “A Lawyer and Technologist Walk Into a Northwestern Lecture Hall“, Strom describes his interview after the class in which he quoted Kris’ “something horrible” observation to Jeroen Plink, a lawyer who’s CEO of Clifford Chance Applied Solutions. Mr. Plink replied:
“Law firms no longer can get away with just working on the basis of the billable hour. And I think competitors will come in. They will attack the firms that are still predominantly doing the billable hour. So yes, I agree with his [Kris’] general sentiment.”
* For instance, AI-based analysis of non-disclosure agreements that are numerous, are frequently presented to the enterprise in the course of its business, and usually relatively simple from a legal standpoint; or high-volume due diligence document reviews for big M&A deals that can be handled by AI-based contract analysis software.