Entrepreneurial Spirit. Independent expertise.

The Courtroom of the Future – Trial by Robot?

Over the course of my career providing litigation related consulting services to attorneys, I have witnessed some of the ways that technology has impacted litigation. Electronic case filing allows for pleadings and motions to be filed from anywhere and, I suspect, has saved many a last minute dash to the courthouse. E-discovery provides for the production and review of thousands of electronic files and emails and document management software scans, indexes and stores documents (eliminating banker’s boxes full of copies).

Technological developments have also benefitted my profession. Database and spreadsheet software allows for the analysis of megabytes of data and other “number crunching.” Video conferencing allows for depositions to be taken with the witness (expert or not) and attorneys in different locations. Scanned and e-mailed expert reports eliminate the need to finish reports in time to print and bind copies to be hand delivered or FedEx’d before 5:00 pm (effectively extending deadlines to midnight).

Trials have changed, too, since the first time I testified; in particular, in the ways that documents and demonstratives are displayed. In fact, there are now consultants dedicated to assisting lawyers with trial graphics and presentations. From animations to recreations to slides to videos, technology provides new ways to convey information to the jury. Jurors using social media have impacted trials and prompted new model jury instructions (see previous GBQ blog posts here and here).

For the most part, these technologies help attorneys, witnesses, juries and judges prepare for, present and decide litigation matters. What if there is (or may be) technology that would eliminate the need for some of these people altogether? What might the courtroom look like in ten or 20 years? I came across two articles that provide a potential glimpse into the future.

Can computers review documents better than people?

Document management software has been used for some time to manage documents and allow for multiple users (including outside experts) access to search, review and categorize documents relevant to that particular user. In large litigation there can be millions of pages of documents that are often reviewed by armies of contract attorneys being paid by the hour. The cost of this type of review has been estimated at more than $1 per page. Costs just for document review can quickly reach into the millions of dollars for a large case.

New technology has started to change this equation. Software – instead of attorneys – can be used to review a population of documents to determine the relevant ones. Although used in only a few litigation cases to date, “predictive coding” (as this technology is known) could change the way that documents are reviewed – especially in large cases. In the case described in the article that I reviewed, the cost of this electronic document review was estimated to be less than half of the cost of a typical, attorney-driven document review.

Although used sparingly to date, the use of predictive coding is expected to grow after court rulings in the past year have approved the use of this type of software in civil cases. Despite its potential for reducing litigation costs, there is still some doubt that even the most sophisticated software can be “taught” to identify all of the nuances of the English language that can sometimes make the difference between a document being relevant or not.

Furthermore, cases with the most documents are often the most critical – with significant financial ramifications — to the parties involved. Each litigant and their counsel will have to weigh the potential cost savings against the risk that a computer would fail to identify a key document in a “bet the company” dispute. Attorneys may find themselves providing not just legal advice but advice about the pros and cons of selecting the right software to perform the document review. Financial experts will have to assess whether the computer was “taught” the right parameters to identify all of the documents relevant to their damages, accounting or financial analysis. Given all of this potential uncertainty, I am willing to venture that humans will continue to play a significant role in reviewing documents in most cases for the foreseeable future.

Can a computer be your peer?

The second article is titled “Could Robots Replace Jurors?” As is acknowledged in the article, this seems particularly unlikely (if for no other reason than it would seem to this non-attorney to be unconstitutional [unless the definition of “peer” changes] in the US). The article actually describes software that is under development to identify false or deceptive statements and designed to be used as an investigative tool. Computers were “fed” transcripts of court room testimony that was subsequently determined to be false. The computers were then “taught” to identify patterns in the testimony associated with the falsehoods. At the moment, the computer is better at identifying true statements than false statements.

While it seems far-fetched to think about the possibility of robots replacing jurors, there are some potential benefits – no more lame excuses to avoid jury duty, no more jurors sleeping during testimony and no more successful appeals to the emotions of the jury rather than the law. Despite these potential benefits, the current state of this technology suggests that our trial by jury system – potential flaws and all – will be with us for many years to come.

Would we even need to have trials?

Perhaps the most interesting implications for the future of litigation would be a combination of these two ideas. If a computer could accurately review documents and determine the relevant information contained in those documents, couldn’t the computer also be “taught” to crunch the numbers to determine damages, apply the law to what it learned from the documents and render a verdict without the need for a trial and jury deliberations? Setting aside the impact on all of the people employed by the legal system (including consultants like me), is it conceivable that a computer with access to all of the documents, transcripts, computational knowledge and the law could render a decision more efficiently than our current system? It seems unlikely, but personally, I hope that I am retired long before that day arrives.


  • Keith Hock
  • Director, Financial Advisory Services
  • (513) 252-0223