Tech Disruption and the Future Role of Lawyers (part 3)


By Phoebe Churches

This is the third post in a series on this topic. If you missed the first two, you can catch up on the first one here and part 2 here. This post looks at

SkyNet, Tech Singularity and the End of Lawyers

“I don’t blame you,” said Marvin and counted five hundred and ninety-seven thousand million sheep before falling asleep again a second later.[1]

So far this discussion has focused on the contracting role of lawyers, and the indications that this contraction will continue apace. Now I want to look at how close this event horizon might be. There are differing views on the immediacy of impacts of automation and technological change on the legal sector. From one side, a headline screams ‘Robots replacing lawyers a ‘near certainty’,[2] and a Deloitte Insight report claims ‘that 39% of jobs (114,000) in the legal sector stand to be automated in the longer term as the profession feels the impact of more “radical changes”’.[3]

On the other side experts at the Massachusetts Institute of Technology tell us that ‘[a]utomation is advancing, but we are still far from the day when machines can do complex physical and mental tasks that are easily and cheaply done by humans’.[4] Similarly, an attendee at the CodeX Future Law Conference at Stanford Law School in May this year recounts much discussion teasing out the difference between ‘what’s real and what’s marketing buzz in artificial intelligence’.[5] A quick survey of Twitter dialogue hash tagged #futurelaw discloses general agreement that the role of Artificial Intelligence (AI) for the foreseeable future will be to assist lawyers rather than replace them. There is some consensus that ‘the notion of the robot attorney is pretty much hype and we still have a long way to go to realize the potential of a fully AI attorney’.[6]

The Tech Paradox

One of the paradoxes of technology is that “simplification complicates”, that is, the more technology you throw at the problem in order to simplify it, the more complex it actually becomes.[7]

While automating transactional processes and other areas of simple decision making has already been a particularly effective tech intervention, complex decision-making processes are still not especially well suited to automation.[8] Moreover, some claims about technology really need to be properly put to the proof. For example, the facial analysis software which can purportedly pick criminals and terrorists by their visage sounds a bit too much like phrenology for comfort.[9] Similarly, apparently accurate predictions can prove to be a fluke. The potential for complex systems to rely on the wrong data is a stark reminder of the shortcomings of current AI. One anecdote which provides a good example is the AI system designed to detect the difference between dogs and wolves. After ‘training’ the system, it had a hit rate of close to 100%. Unfortunately, the system was simply detecting the presence of snow as a common element in all of the wolf photos, where the dog pictures featured none.[10] There are also troubling possibilities brought about by operator error and/or bugs introduced during coding of systems. These are not trivial concerns in the context of legal processes, and it may be a long wait for ‘the arrival of ultra-reliable and verifiably crash-proof code … a holy grail in the development of increasingly complex systems’.[11]

Ultimately though, these issues are about the rate of progress, rather than the inevitability of change. The writing on the wall is clear. Humans are no match for machine intelligence and efficiency in an enormous range of tasks. For example in the 80s and 90s, a large chemical company ran work done by its in-house legal staff through new data-mining software and found a human accuracy rate of only 60%. That is a lot of money spent on salaries for outcomes only ‘slightly better than a coin toss’.[12] It is indisputable that data-driven models can help make better legal decisions; yet, for the moment at least, and ‘for the appropriate tasks, the age of quantitative legal prediction is a mixture of humans and machines working together to outperform either working in isolation. The equation is simple: Humans + Machines > Humans or Machines’.[13]

The Regulatory Challenge

There is another factor limiting the speed of development in the sector: the full impact of rapid technological development continues to be throttled by slow regulatory change. Current regulatory barriers compromise services to consumers on both ends of the spectrum – on one end, entry barriers have created the monopoly which has facilitated a false market in legal services, and has limited competition from outside the sector which might otherwise weed out slapdash or underperforming firms. On the other end of the continuum, services in the unregulated space can enter the market unimpeded, providing all manner of products and services to unwary consumers with relative impunity.

While America is yet to reform its regulatory framework which enacts substantial barriers to entry and practise, the UK and to a lesser extent Australia have undertaken reforms to allow alternative business structures (ABSs). However, these reforms will need to go further as technology increasingly pushes the existing boundaries of regulation.[14] The Legal Services Board and the Solicitor’s Regulation Authority in the UK are actively promoting extensive regulatory reform to accommodate increased segmentation in the legal services market.[15] In Australia, incorporated legal practice and multi-disciplinary partnerships have been permitted for some time, however these models are still tightly confined by the regulatory framework.

The need to tread a careful line between freeing up the sector to embrace change, and protecting clients and society generally accounts in part for the sluggish rate of change to regulation. Witness the story of Justin Wyrick Jr who, in 2000 became the most asked for legal expert on ‘Justin’, as it turns out, was in fact Markus Arnold, a 15-year-old secondary student who had never opened a law book in his life.[16] Mr Arnold was not prosecuted, to the American Bar Association’s abject horror, however his efforts are a pretty clear indication that consumers need some protection. Free online services are not currently regulated by consumer laws, so minimally we need some accreditation based regulation as assurance so the community can have some faith in what its (not) paying for.

At the other end of the spectrum, regulation unnecessarily interferes with potential improvements to the accessibility of the legal system. For example, in the US State of Florida, Rosemary Furman assisted people wanting a divorce by preparing and filing the necessary legal forms for $50.[17]  Ms Furman had previously done this work as a legal secretary under the supervision of an attorney who charged $300 to complete the same work. She thought the cost of filing for divorce was unconscionable, particularly for women unable to afford to leave violent relationships. [18] Unfortunately Furman was a victim of her own success, because her business attracted the attention of the Florida regulators who sentenced her to 120 days in gaol for her efforts. It was only by intervention of the Governor that she did not actually serve time.

Ultimately in the context of a disaggregated sector, regulators need to find ways to protect the interests of clients, but without erecting unnecessary barriers to entry, and constricting innovation. This has not proven a problem for the legal work increasingly undertaken by accountants and conveyancers. It is difficult to see why there is any barrier (other than the self-interest of lawyers themselves) to employing the same flexibility to encompass the increased segmentation of the legal sector. Mayson makes the case for ‘maintaining sector-specific regulation, rather than leaving legal services to be covered only by general consumer and competition protection’.[19] Where the stakes are particularly high for clients, such as ‘the potential for irreversible loss, misuse of clients’ funds, or abuse of a privileged relationship’,[20] there needs to be specific consumer protection, above and beyond the current regime. One way or another, regulators must recognise that the unbundling of legal work has at once opened up opportunities to address unmet legal need, and a potential space for the uninitiated and unannointed to wreak havoc.

Previously: The Contracting Role of Lawyers | Next Time: The New Frontier

[1] Douglas Adams, The Hitchhiker’s Guide to the Galaxy (1982).

[2] Miklos Bolza, ‘Robots replacing lawyers a “near certainty”’ (22 Feb 2016) Australasian Lawyer.

[3] Deloitte, ‘Developing Legal Talent: stepping into the future law firm’, Insight Report (February 2016).

[4] Timothy Aeppel, ‘Be Calm, Robots Aren’t About to Take Your Job, MIT Economist Says’ The Wall Street Journal (25 February 2015).

[5] Miguel Willis, ‘Robot Lawyers: Kill Law Jobs or Augment Expertise?’ (24 May 2016) The Innovative Law Student .

[6] Ibid.

[7] Frank McKenna, ‘In the zone: Is technology helping or hindering lawyers’ decision making?’ (September 2013) LexisNexis Australia Discussion Paper.

[8] Ibid.

[9] Debra Cassens Weiss, ‘Company claims its technology can pick out criminals by facial analysis’ American Bar Association Journal (24 May 2016).

[10] Ibid.

[11] International Legal Technology Association, ‘Legal Technology Future Horizons – Strategic Imperatives for the Law Firm of the Future’ (Report, 2014).

[12] Bregman, above n 18, 12.

[13] Daniel Martin Katz, ‘Quantitative Legal Prediction—Or—How I Learned to Stop Worrying and Start Preparing for The Data-Driven Future of the Legal Services Industry’ (2013) 62 Emory Law Journal 909, 929.

[14] Steven Mark &Tahlia Gordon, ‘Innovations in Regulation—Responding to a Changing Legal Services Market; (2009) 22 The Georgetown Journal of Legal Ethics 501.

[15] See e.g. Legal Services Board, A blueprint for reforming legal services regulation (September 2013); and Solicitors Regulation Authority, SRA Regulatory Reform Programme Improving Regulation: proportionate and targeted measures (April 2015).

[16] Clifford Winston, Robert W. Crandall and Vikram Maheshri, First Thing We Do, Let’s Deregulate All The Lawyers (2011).

[17] George C. Leef, ‘The Case for a Free Market in Legal Services’(October, 1998) Policy Analysis No. 322 – the CATO Institute, 1, 2.

[18] Ibid.

[19] Stephen Mayson, ‘Beyond the Legal Services Act’ (27 July 2015).

[20] Ibid.