Lawyers are turning to machine learning to ease caseloads

Posted Sep 4, 2017 by Tim Sandle
Lawyers are making use of machine learning and artificial intelligence platforms in order to assess past cases and to make predictions about the potential success or failure of new cases.
Scales of Justice
Scales of Justice
Easy Stock Photos
As with many other services, the legal profession is undergoing digital transformation. Established firms are facing competition from aggressive startups running digital only platforms. This disruption is requiring legacy firms to adapt.
Meeting the challenge of digital transformation is not only necessary to remain competitive in terms of offering alternative channels, the use of digital technology can also assist with the amount of work involved with the legal process. As firms seek to adapt, machine learning is being adopted more and more in order to help companies grow more efficient and produce greater value from their processes.
Assessing case history is an important part of the legal process. Other voluminous amounts of data also need to be processed. This includes the client’s history, plus case notes, briefings, testimonies and other data. If this data is digitized, then searching and interpreting it is made easier. To assist legal teams with this data review process artificial intelligence platforms are being adopted. In addition, intelligent machines can be used to assess decisions made by judges — to work out how case law is set and what the decision of a particular judge is likely to be.
According to a review by Forbes, the use of machines to assess legal cases can reduce the hundreds of hours required to trawl through case notes considerably. In addition, machines can ‘learn’ areas of potential interest, such as unusual patterns, and flag these to the legal team. In addition, the firm LexisNexis DiscoveryIQ estimates using analytics can save 70 percent of the expenditures involved in legal reviews. This happens thorough machine learning algorithms constructing computer models that can interpret complex legal text by detecting patterns and inferring rules from data.
As assessed by the company Innovation Enterprise, the two largest companies in legal data-driven research are LexisNexis and Westlaw. These operations possess databases containing vast numbers of case details.
Databases, however, are not machine learning platforms. A system called Brainspace has begun applying analytically driven tools to unstructured data. One such application is using this type of machine to trawl through millions of unsorted, unstructured documents. The device has been tested in relation to the documents from the Enron Scandal. The original process of documentation review took lawyers months to go through; whereas, in a simulation, Brainspace completed the same feat in under an hour.
An alternative system has been used by JP Morgan to assess financial law. The program is called COIN (Contract Intelligence) and it assists the firm with the taxing task of interpreting commercial-loan agreements. The software has, as the Independent reports, saved hundreds of hours each month in personnel time.
Another example of technology shaping the legal sector, where the use of human lawyers can be reduced, is Legal Robot. This system runs automated contract reviews, to assist businesses in understanding complex legal language of contracts and to spot problems with contracts before they are signed.
These three examples show the different types of machine learning technology that can be used by law firms and the different drivers for the adoption of the technology. Lawtech remains at an early stage (say compared with Fintech); however, as it takes off, the disruption is likely to be considerable.