- AI won’t replace a lawyer’s judgment.
- AI will drive labor-saving technology to perform law’s “manual” tasks.
Jeffrey Carr put this best:
— Jeffrey Carr (@CarrNext) March 22, 2017
Jeff Carr is that rare general counsel who reduced total legal costs and proactively headed off liability before it arose — instead of the steady, single digit increases in legal budgets that are the norm.
At FMC Technologies between 2002 and 2013 he went from $14.3 million to $9.5 million between 2002 and 2013 — amid a 4X increase in revenues!
I am partial to Jeff Carr’s assessments about the practical effect of AI on business law because he has a background that almost no other in-house lawyer shares: Carr has run his own company and managed to a P&L.
So … AI won’t replace a lawyer’s judgment.
- Should a pharmaceutical company sue a competitor for apparent infringement of its patent?
- Should a trucking company start to carry frieght that brings its operations under the Hazardous Materials Regulation?
- Can the business fire an employee based on the progressive discipline file that its human resources group has compiled?
- Where the debt limitation covenant in a lending agreement with the bank is ambiguous as whether or not a particular form of financing counts against the limit — should we risk being sued by the bank if we think that we need that financing?
Judgment calls rarely have just one reasonable answer.
And they only start with the technicalities of what legal institutions demand — technicalities alone don’t provide the answer — for two reasons:
First, the demands of legal institutions are often ambiguous, subjective, or erratic. It’s rarely certain what a court or regulatory agency will actually require of your business based on a judicial opinion, statute, or regulation.
Second — in a context where technicalities can be ambiguous, subjective, or erratic — it’s vital to understand your company’s goals and its industry in which you compete when you’re navigating such uncertainty.
As a general manager I encountered too many lawyers weak in their grasp of the company’s goals or its industry. Worse — lawyers are trained to be risk-averse — and many take that training too much to heart.
Excessively risk-averse attorneys tend to resolve doubt in favor of inaction.
Judgment calls in law call for considered decisions where the law itself is open to more than one interpretation.
So … don’t look for AI to replace a lawyer’s judgment.
And … AI will drive labor-saving technology for law’s “manual” tasks.
Law’s “manual” tasks consist of legal research and document review as described in Part II of this series:
- In a bankruptcy case: Find me all of the cases decided in a particular federal district and its appellate circuit that construe the standards by which the debtor’s fraud would disallow its discharge in bankruptcy (e.g., Ross Intelligence)?
- In a Title VII sex discrimination case: Look at e-mails sent or received by the plaintiff’s manager and by anyone in the human resources group that discuss the plaintiff — seeking words that may indicate gender-based animus (e.g., Brainspace or NextLP).
- Tax regulations: Review all equipment leases to which a client company is a party and compare their terms to a new federal income tax regulation — asking which are compliant and which will need to be changed (Ernst & Young)?
- Due diligence: Here are the contracts to which the target company in an M&A deal is party — identify those with a change-of-control provision that might require novation or some sort of permission by the counter-party (Axiom’s Contract Intelligence Platform).
Two observations about AI and “manual” legal tasks:
First, these are the tasks that have traditionally been done by less experienced, younger lawyers at hundreds of dollars per hour.
And — as Jeff Carr’s tweet indicates — these labor-intensive activities are at the heart of the legal industry’s (mostly wasteful) hourly billing and pursuit of leverage business model.
Second, when people do mind-numbing work — like reviewing hundreds of emails for wording that indicates gender-based animus — mistakes are likely.
AI offers greater accuracy in “manual” tasks as well as lower labor costs.