The technology freight train is barreling down the track; for those law firms and lawyers hobbled and bound by an inability or unwillingness to innovate, the collision will be a messy one indeed.
One innovation rapidly gaining traction is the use of legal expert systems by law firms and clients alike to provide quick answers to repetitive legal questions. Legal expert systems are software applications that enable a user to solve legal decision making tasks by answering questions posed by the software. The decision tree upon followed by the expert system is created by one or more subject matter experts. Ideally, the interaction with the expert system should simulate the give and take the user could expect if engaging a subject matter expert to analyze the problem. Once the expert system has sufficient information, it provides its conclusions to the user. One familiar example of a widely used expert system is TurboTax. You answer its questions, it applies the rules behind the scenes and then it spits out a completed tax return and related documents.
Expert systems have been around for decades, but with the continued exponential advancements in computer processing power and memory, these systems have drastically improved. It is only a matter of time until in-house counsel will have ready access to expert systems that can answer many routine legal questions without the need to consult outside counsel. Perhaps such systems will be provided on a subscription or other fee basis by law firms or as an added benefit of retaining the law firm for non-routine matters. It is more likely, however, these systems will be provided by the likes of IBM or other technology companies playing the role of disrupters. The success of IBM’s Watson system at defeating the best two Jeopardy players in the World is well known. IBM is investing over a Billion dollars in transforming Watson into a legal and medical expert. Have no doubt Watson and similar systems will replace many lawyers—and it will not be long before lawyers start noticing.
For “fun” I recently created a simple expert system using Exsys Corvid Core, an application for the Apple Mac. (App Store $149.99). Corvid Core is relatively easy to use, even without prior programming experience. The system I created helps the user determine whether a putative class action filed in state court is removable to federal court under the Class Action Fairness Act (“CAFA”). The decision tree upon which the system was based was one I created several years ago and is the title photograph for this article. To use Corvid Core you construct a similar decision tree, which looks like this:
While Exsys claims it is possible to make this expert system active on a website or intranet, I have not quite figured out how to do that yet. For now, I can only show you a screen recording demo of this simple expert system in action.
The important take away is that an expert system can be created for any problem that can be resolved by answering questions on a decision tree. This would encompass most routine legal questions. One company, Neota Logic is helping law firms to catch the expert systems train instead of being run over by it. However, they are not putting all their eggs in one basket and are also selling the same expert systems technology to corporate law departments. They also suggest that companies request their counsel to create expert systems to answer routine legal questions:
“Many elements of a large company do not benefit from, or even seek, legal advice because obtaining advice is costly and time-consuming. Neota Logic applications meet this latent demand for legal advice by delivering interactive applications on the company intranet that can be consulted by managers and other business people to get situation-specific guidance around the clock and across the globe.”
All indications are that the use of legal expert systems is likely to increase dramatically in the next few years. Smart law firms will either use these tools to bring additional value to their clients or risk losing clients to those that will.