Lawyers are not usually known for being enthusiastic about change, but when it comes to technology, they may not have a choice. A recent study estimated that 23% of legal activities could already be automated. Another report proposed that this number may actually be 39%. In his often-quoted book Tomorrow’s Lawyers, Richard Susskind predicts that in the next two decades, the way in which lawyers work will drastically change and traditional businesses that fail to adapt will undoubtedly not survive. These predictions may be concerning because, taking the U.S. as an example, the number of lawyers over the last ten years has increased twice as fast as the population growth, reaching around one lawyer for every 245 Americans. Should lawyers be worried that computers will replace their jobs, and should law firms be concerned about their current business model?
By 2011, news had already appeared that computers, being able to identify relevant words and phrases, were displacing lawyers from discovery practices. Today, companies such as Blackstone Discovery, Disco, Discovia and Relativity, to name a few, provide e-discovery services that greatly lower the time and costs of searching through documents. JP Morgan’s COIN (for Contract Intelligence), one of the platforms that reviews contracts, has reportedly cut the time of reviewing the company’s commercial-law agreements from 360,000 hours per year to a simple matter of seconds.
For years, databases such as LexisNexis and WestLaw have allowed easy access to caselaw and other materials that are essential for attorneys to give legal advice. Other tools are being developed to help even further. IBM’s Ross, for instance, uses Watson – IBM’s artificial intelligence – to understand specific inquiries by the user (e.g., “what is the standard for gross negligence in New York after 2004?”), after which it would search the data-bases, identify patterns and present the answer with the most relevant cases in the given jurisdiction and timespan.
At first, these technologies may not seem revolutionary. Arguably, they belong to the category of sustaining technology rather than the so-called disruptive technology (the former being those that foster improved performance and the latter those that fundamentally challenge and change the functioning of a given business). Nevertheless, the impact that such technology can have on a lawyer’s activity is not to be overlooked.
The first draft of an agreement could be ready in mere minutes rather than hours. Document review could require a single license for a software rather than a whole team of lawyers. Instead of having an associate perform hours of case law research, an AI platform could quickly provide the current state of the law in a given jurisdiction in minutes. All these and other tasks that are an essential part of a lawyer’s service could easily be done by computers in less time, at a lower cost to the client and possibly with a greater level of accuracy. So what does that mean for the legal profession?
Will Technology Replace the Work of Lawyers?
Garry Kasparov’s loss to IBM’s Deep Blue in a chess match still lingers as a decisive moment in history, as machine intelligence reached the point of overcoming human intellect. However, Deep Blue did not have “intelligence”. It merely searched its database and responded to preprogramed rules in order to find a solution faster than a human being would be able to. Actual intelligence – the kind that could, allegedly, replace lawyers – requires more.
Perhaps the most famous test was proposed by Alan Turin’s “Computing Machinery and Intelligence”, in which a computer could be deemed intelligent if a person could not determine if it was speaking to a machine or another person. However, the “imitation game” does not test actual “intelligence”, but only the simulation of intelligence. For instance, a person who does not speak Chinese can, with the aid of an English-Chinese dictionary, translate a message and simulate knowledge of Chinese. The leap to artificial intelligence requires that the machine be capable of not only responding to pre-determined rules, but of recognizing patterns, learning from past experiences, analyzing the problem structure, planning how to tackle the issues and finally inducing inferences from past analysis in order to solve current problems (see Marvin Minsky’s seminal 1961 work “Steps Towards Artificial Intelligence”).
While software that uses artificial intelligence in a legal environment does exist today, it is highly unlikely that computers will completely replace lawyers’ jobs. A machine may be able to gather all the information about contract law and the precedents in a given jurisdiction and it may even identify the arguments and counter-arguments that have been more persuasive in a given court. However, identifying a group of cases in which a given concept (e.g., good faith) was applied does not mean that the machine has an actual understanding of the underlying concept. As in the imitation game, this may only simulate knowledge.
That is not to say that the products in the market today don’t evaluate – and most of the time correctly – legal documents and relevant case law. A machine, however, may have difficulty determining whether a precedent related to the sales of apples is applicable to the sales of oranges or the sales of imported cars. It is doubtful that a machine faced with the classic “No vehicles in the park” problem would be able to identify if it applies to bicycles, wheel chairs and strollers, or if there may be an exception for ambulances in emergency situations, or if a military truck set on a pedestal as a memorial would violate this rule.
Many examples could be given of similar situations. The bottom line is that technology, at least for now, is not going to replace lawyers’ tasks such as writing legal briefs and pleadings, advising clients and negotiating with counter parties. It will, nevertheless, deeply impact the way in which lawyers and firms conduct their businesses.
Impact of Technology on the Practice of Law
In 2017, Florida became the first State in the U.S. to add technology training as part of the continuing legal education requirement. In 2018, the American Bar Association amended its comments to Model Rule 1.1 related to competence, adding that “to maintain the requisite of knowledge and skill, a lawyer should keep abreast of changes in the law and its practices, including the benefits and risks associated with relevant technology […]”. It may not be far-fetched to imagine that in a few years, not using technology may lead to malpractice liability.
So how exactly will technology change legal work? Some of the most obvious implications have already been mentioned.
Document review can be done by software in a much faster, and possibly more accurate way. For example, lawyers could “teach” the machines about what they are looking for – e.g., specific keywords or communication between Mr. X and Mr. Y related to a given topic – and the computer would flag the relevant documents. This decreases costs and makes the lawyer’s work much more efficient, as she does not need to go over bundles of irrelevant documents. It is noteworthy that at least in the Southern District of New York, the use of AI-powered software for discovery has been widely accepted at least since 2012 (See Da Silva Moore v. Publicis Groupe).
Analyzing and drafting contracts will also be made much easier. Companies sign new agreements on a regular basis which are usually reviewed by lawyers. The massive number of hours required to do this job can lead to a bottleneck of backlogged contracts that jeopardize both the company whose deals may be delayed, and the lawyer, who cannot do other work. Imagine, for instance, the overwhelming job that due diligence requires. As with J.P. Morgan’s COIN, computers can analyze the data and provide summary charts, or with Allen & Overy’s and Deloitte’s MarginMatrix, present initial drafts of agreements based on the applicable regulations in the relevant market, cutting a considerable number of hours.
Legal research is also a clear area that is being affected by technology. If one cannot imagine doing research without databases such as LexisNexis and WestLaw, a few years from now the same may be said about more advanced research tools, such as IBM’s Ross. Again, this will decrease the number of hours attorneys need to dedicate to the job, ultimately decreasing the cost for clients.
Online services are already common in other fields and will increasingly be so in law. Users will likely turn to the Internet rather than lawyers to produce standard documents such as wills, landlord and tenant agreements and so on. Litigation will also be affected by online services in the coming years. Beyond the obvious digitalization of procedures that used to be done in paper (so that briefs, documents and decisions can be filed online and viewed instantly by any user), online dispute resolution mechanisms are starting to appear that will reduce the amount of cases that are actually brought before a court. The European Union, for instance, has created a platform allowing for consumer claims to be resolved by an independent body without the involvement of national courts.
Predicting results is one other example of the many ways that technology will affect the work of lawyers. With an input of several cases by a given court, algorithms can predict, with more than 70% accuracy, the outcome of a given dispute. This can largely aid lawyers in deciding whether to take on a case, and, if so, the fee structure to be used. Similar software could help in-house counsel in assessing which matters they want to handle by themselves, which they want to settle and which they would require advice from external counsel.
Alternative Business Structures are also one of the main changes caused by technology. Law firms today use virtually the same business model as they did a century ago. As shown by the examples above, new technologies will drastically reduce the need for attorneys – especially junior lawyers – to do tasks such as document review, contract drafting and legal research. Those services could be provided faster and with lower costs by a machine. This deeply affect the pyramid-structure used by law firms, with a considerable amount of young associates at the bottom and a decreasing amount of lawyers in the upper levels.
With that in mind, the UK set out to revise its regulatory framework for legal services, enacting new legislation that allows non-lawyers to own and run legal businesses. (Similar practices were also adopted by Germany in 1999 and British Columbia in 2010). This allows outside investors to inject capital into legal businesses and become co-owners of law firms. That also means that multidisciplinary practices – such as accountants, business advisors and lawyers – can all be provided by the same enterprise.
“The first thing we do, let’s kill all the lawyers” was proposed by Shakespeare in one of his plays. It may seem, at a first glance, that technological developments may do exactly that to the legal profession. But a closer look reveals that technology will not replace lawyers so much as change (perhaps considerably) the way in which they conduct their businesses.
Automated solutions will drastically decrease the number of hours needed to complete tasks such as legal research, document review and contract drafting. If that may appear as a benefit for small firms – which can now take on more cases without being overwhelmed by the volume – for bigger law firms that means narrowing the bottom of the pyramid and reviewing their business models. There is no stopping progress and law firms must be ready to adapt or otherwise face the risk of being left behind.
Marcel Carvalho Engholm Cardoso is an LL.M. Candidate in the International Business Regulation, Litigation and Arbitration Program at New York University Law School. He is a graduate of Pontifical Catholic University of Sao Paulo and holds a Master’s Degree in Civil Procedure from the University of Sao Paulo. He practices litigation and arbitration in Sao Paulo, Brazil.