Some Thoughts on Legal Tech

Starting new posts is always awkward for me, as I feel the need to provide a number of disclaimers. In that spirit, I should note that I’ve never worked as a lawyer, nor have I worked in legal tech. However, I do have a law degree and a master’s in law and finance, and I also am a programmer. This also is my website, so I can write whatever I want. Decide for yourself if it is useful.

Of course, the term “legal tech” is incredibly broad, and in this post I make no attempt to cover the area in its entirety. Instead, I want to focus on how technology could transform the commercial law market. Corporations and high net worth individuals are going to continue to spend large sums of money on legal services; I am interested in how legal tech will be used to provide those services, and how it might alter the competitive landscape of the legal market.

My general hypothesis is that technology has the potential to radically transform the legal industry, but that it is very much uncertain whether this will occur any time soon. There are a number of impediments to the widespread adoption of legal tech that have nothing to do with technical feasibility.

Types of Legal Tech

The legal tech that is relevant to this post broadly can be divided into two groups. First, there is “efficiency enhancing” legal tech: this type of tech makes lawyers’ jobs easier and often saves them time as well. Ordinarily, it is the work of junior lawyers that is being reduced. I recently talked to a first year associate in the US, and she said (roughly):

Image Did Not Load

Clearly some examples of this type of legal tech are already common at law firms, such as document comparison tools, legal research services like Westlaw, etc. However, there is quite a lot that has yet to be developed that would further drive efficiency.

Second, there is “quality enhancing” legal tech: this type of tech does not necessarily reduce hours worked, but simply makes lawyers better at what they do. It actually results in the client getting a better service. Of course, much of legal tech that fits into the first type can also fit into the second type; I can’t imagine that lawyers had quite as comprehensive of a search before legal research services like Westlaw came around, and it is unlikely that human eyes were quite as good as software at spotting differences between documents. However, when I refer to this second type of legal tech, I am referring to legal tech that has quality enhancement at its core.

This second type of legal tech is exceptionally rare at the moment. However, if you’re the type of lawyer who worries about tail risk, this is the type of legal tech that you should be thinking about. It is entirely possible that legal tech will produce large improvements in the quality of services offered, rendering any lawyer who lacks access to such tech no different from a butcher without a knife.

From a technical perspective, I am quite optimistic about both types of legal tech. I don’t think we need any major breakthroughs or developments in machine learning or neural networks in order for these products to exist. But someone has to build and use these products, and that is where I think there is room for uncertainty.

The Firm’s Perspective of Type 1: Balancing Efficiency and Sustainability

Law firms, just like all firms in a market economy, have an interest in cutting costs. Some lawyers express this in terms of competitive pressure: their clients are price-sensitive, and they therefore have to find ways to lower costs while maintaining their profits. Others would just say reducing costs means increased profits for the partners. Regardless of who captures the benefit of the cost savings, the desire to cut costs is real. The first type of legal tech mentioned above therefore seems like a no-brainer.

However, I am not convinced it is quite that simple. There is a fear among some law partners that automating the work of junior lawyers will be harmful to their training. While this might sound like your high school math teacher who told you that you won’t always have a calculator with you, or your grandfather who complains that you don’t know the value of hard work, I think there’s something to this idea.

Software might render it unnecessary for junior lawyers to spend hours sifting through contracts, reading the terms of hundreds of contracts in full. And while the skill of skimming through contracts might never again be needed, the knowledge that comes from doing so is nonetheless useful. We might automate the search process, but the senior lawyers will still need to know about the clauses and the variation that appears in the wording. The tedious work that is now being automated might render junior lawyers inadequately prepared for the work of a senior lawyer.

This causes problems for the sustainability of the firm, as some junior lawyers will one day be partner. If the junior lawyers are failing to learn the necessary skills, a young partner today will be understandably worried. Of course, I don’t mean to suggest that this is an insurmountable challenge for legal tech; new methods can be developed to train young lawyers, after all, and it is likely that the new methods will be more efficient than the old methods. But if these methods are not developed alongside the adoption of legal tech, there is going to be a problem.

The Firm’s Perspective of Type 2: Who Cares About the Future?

In contrast to the above problem where a firm might be concerned about its future sustainability, firms are skeptical of the second type of legal tech because the benefits might not accrue for many years, at which point many of the current partners will be gone. Developing and implementing legal tech within a firm that truly makes a lawyer “better” takes more time and effort, and the gains might not be realized for years. Law firms, as partnerships, do not care about investments that will pay off in the future; they care about profits today.

An easy way to square these two problems, one in which the firm is worried about long-term sustainability, the other in which the firm is unconcerned with future profits, is to consider the roles of different partners. In the first case, the older partners would prefer to adopt legal tech because there is an immediate payoff, whereas the younger partners would have concerns. In the second case, the younger partners would prefer to adopt legal tech because there is a future payoff, whereas the older partners would not be interested. I am sure it is more complicated than this, with many partners never thinking in these terms, but these are the motivations that allow a firm to hold both concerns simultaneously.

It also seems that most lawyers doubt that this second type of legal tech is even possible, or at least haven’t thought much about it. This is partly due to arrogance I’m sure, in that a tax lawyer might not be so receptive to the idea that a computer would be better at designing tax mitigation strategies. But it is also because lawyers don’t get paid to think about these things, and a lawyer will probably feel he or she is doing the job exactly as it always has been done, and doing it well. This isn’t the industry for innovation.

Making Legal Tech Work

The first type of legal tech is inevitable, and the development of such technology will continue its steady march forward. The progress might be far slower than it needs to be, as has been the case to date, but the pressure for efficiency will ultimately win. The second type of legal tech is far less certain. Realizing the value of this technology within a firm is a challenging task, and clients aren’t pounding on law firms’ doors demanding it either. I am sure the value of the second type of legal tech will be realized one day, but that may be a long time from now.

But my personal view is that partners in law firms should start exploring this path. Maybe no other firm will do it, and these legal tech products will only be developed by tech companies (insofar as they are developed at all). But what if a competitor firm builds the technology itself? What if a competitor firm owns technology that makes your lawyers into butchers without knives? I am not saying that is likely, but I do think it is possible.

Of course, saying “I would go for it” doesn’t overcome the challenges discussed above. In relation to the problems with the first type of legal tech, I think the solution is to focus equally on ensuring that lawyers whose work is being automated receive proper training nonetheless. If the problem is that they no longer are exposed to enough contracts, for example, have the lawyers use the technology to find the variations in wording across clauses and the frequency of each. Hell, teach the lawyers how to run a regression to discover if any trends can be found as to when certain wording is used.

In relation to the problems with the second type of legal tech, the real challenge is in getting the programmers to add value from the start. Partners will not tolerate large investments made if the benefit only accrues in the future, so a firm needs to ensure the cost is at least somewhat justified in the present. Products should be developed incrementally so that they can be useful as soon as possible, and the programmers should be integrated into the firm so that they can begin to add value on an ad hoc basis from the start.

This post is getting long and starting to sound like a pipe dream, so I will end it here. I don’t know whether any firm will attempt to follow a strategy of developing and adopting legal tech, or whether the advice I gave above would even work. However, any firm that isn’t thinking seriously about legal tech is betting either that I am wrong and legal tech’s potential is limited, or that no other firm will take legal tech seriously either. And I am not sure whether that is a good bet.