The Costs and Benefits of Automating Legal Practice
Law.com recently provided the details of a report undertaken by a commercial property firm in London that found 89% of London law firms “either use AI already or plan to do so in future.” As law firms bring more automation into their practices via limited forms of AI, it is useful to briefly assess the costs and benefits.
First the benefits.
Ross Intelligence, an IBM product aiming high-quality AI at legal practices, has recently argued that the proven advantages of automating legal research and document production are very close to those predicted in the summer of 2017 by the ABA blog ‘Law Technology Today’. The advantages listed by LTT are (quoting nearly verbatim):
Allowing earlier (and more accurate) risk assessment,
Producing higher-quality work,
Improving the organizational and logical structure of firms,
Enhancing creative analysis and identification of persuasive precedents
Reducing attorney stress and frustration, and
Improving client relations.
Of these points, perhaps 3, 5 and 7 beg closer attention. As you might expect, points 5 and 7 go to the fact that lawyers are expected to have more time to do what humans do best – be creative and interpersonal – with machines performing the repetitive tasks. With machines undertaking more of the routine work, lawyers can spend more time with clients and have more energy to develop creative solutions. Point 3, however, is likely not what you’d expect: “produces higher-quality work” does not mean that machines free up lawyers to aim at high-end output. It actually means that computers are better than humans at research and drafting legal documents: “Work produced by intelligent software—which doesn’t get tired, bored, or distracted—can be truly error-free…. Such software can also ensure that language is applied consistently, no matter how many attorneys had a hand in the drafting.” This has been proven in human v machine competitions, even with the relatively rudimentary products now being sold.
IBM’s Ross team adds to the above the very important and relatively obvious factor that “investing in AI leads to higher revenue.” Following sunk costs for purchase and fixed costs for maintenance, use of a computer will allow 24 hour/day, 7 day/week labor with a learning curve that goes sharply upward, given that computers need be told something only once, do not forget anything, and can combine the experiences learned from many applications simultaneously into their single driver (perhaps in the cloud). They also can receive software upgrades periodically.
The second point the Ross team makes is more problematic. They predict that “AI is better for the well-being of attorneys,” which is probably intended to mean that lawyers will have more time for the things they like to do. It could mean also that those lawyers who receive the higher revenues from having replaced other attorneys with an AI system will experience that well-being bump.
That brings us to the costs of automation.
Law.com recently reported with regard to London law firms: “Between 43% and 45% of law firms surveyed by commercial property company CBRE said they believed there will be a reduction in headcount at junior and support levels due to firms’ adoption of AI, with 5% of respondents saying they expect cuts of more than 20%.”
The methods of document review and assembly that senior lawyers once gifted their younger colleagues may now be written up as protocol for AI systems that will reproduce the analysis or drafting technique infinitely. In a sense, these older lawyers will be immortalizing themselves and turning knowhow into a stream of future license fees, similar to what happens when a generation of partners converts the partnership into a corporation and takes it public. This could mean a bump of revenue for the persons whose knowhow is automated, but it will also mean that the value passed on to the next generation will be lower.
Brace for the future.
As automation in the legal profession is just now beginning in earnest, it would seem wise for law students to prepare themselves both to jump above routine research and document production into more creative work early in their career and to learn how they can most productively interact with the new technology being adopted by law firms. Someone planning to read cases, sort documents and draft explanatory memoranda for the first few years of practice could meet an unhappy surprise.
David Donald, Hong Kong