Court Automation: Robot Judges on the Way?

It was 30 years ago now that Deep Thought first tied Grandmaster Anthony Miles in a chess tournament. Nine years later, in 1997, Garry Kasparov, the thirteenth world chess champion, played a super match against IBM’s Deep Blue. Four years ago, the legal robot LegalApe 2.8 proved that computers were ready to participate in a high-level simulation of a real legal dispute. Many legal institutions, like some other areas of human activity, have shown potential for algorithmic development, which involves the use of a clear sequential tree of conditions and decisions leading to specific legal conclusions. The text of many legal acts is based on logical constructs of a similar nature: if a party has committed a certain act, there will be a certain consequence, if the party meets certain characteristics, it is entitled to a particular status, and so on and so forth. As a result, the possibility of automation has been explored for a straightforward category of court cases that follows an algorithm prescribed by procedural codes and considers formal circumstances alone when issuing a judicial act. It was cases of precisely this category that were automated for a pilot court project that took place in the Belgorod region. A virtual assistant justice of the peace was developed to analyze incoming applications for court orders from the Federal Tax Service and prepare drafts of court orders based on the established algorithms. Can AI successfully be used to automate and streamline court processes and free up judges to focus on real legal disputes? Might we see a full-fledged autonomous robot judge in the near future? We’ll be looking for the answers to these and other questions in our roundtable discussion.