HOW NOT TO LOSE A DISPUTE: PREDICTIVE JUSTICE
In our country too, more and more attention is being paid to the so-called “Predictive justice” in order to predict the outcome of a judgment based on the analysis of all previous cases using algorithms.
As recalled by Maurizio Carmignani, already in 2004 a group of professors from Washington University had all 628 cases discussed in 2002 by the Supreme Court of the United States examined: the algorithms were able to predict 75% of the outcome of the judgments compared to 59% of a group of experts.
More recently, the Brescia’s Court of Appeal in collaboration with the local University is carrying out a pilot project, the first in Italy, limited for the moment to Labor Law and Company Law, through purposefully create website – a database of decisions issued in the jurisdiction to provide a set of precedents on which to inform future judgments.
By clicking on Company Law, for example, under the heading “price offer below cost” the wording “the competitor company that offers prices below cost is not required to pay compensation”, referring to a decision of the local Court on unfair competition issues , which excludes it.
In this way, the party and the lawyers who intend to initiate a legal action in Court will immediately have indication of the likely outcome.
In the future, the algorithms will be able to examine all the sentences issued in Italy in recent years, for example on this subject of below cost prices. And then provide a forecast on the outcome of the judgment. Thus making article 360 bis c.p.c. that (in an attempt to reduce the more than 25,000 sentences that the Court of Cassation is forced to issue every year) since 2009 declares the appeal inadmissible when the sentence to be quashed decided the legal matter in accordance with the Court’s jurisprudence.
Already now, particularly structured larger law firms have equipped themselves with an internal algorithmic processor which, based on all the company’s past statements, facilitates the lawyers in drafting new ones, with suggestions based on previous experiences.
And then? Will it come to robotic justice, will robots, in the most recurrent and identifiable standard cases, issue a sentence after examining in a second all the precedents on the subject? In chess, man nowadays always loses.