Technology

Artificial intelligence can also be useful for lawyers

Artificial intelligence can also be useful for lawyers

In the last ten years in Italy, banking disputes, that is the lawsuits between banks and their customers, have increased considerably. This happened, in large part, because unpaid loans increased, which in jargon are called “non-performing loans”, or “non-performing loans”: in 2019 they were worth a total of 327 billion euros.

In fact, since the economic crisis of 2008, more and more people or companies have not been able to repay the amount due, whether it is mortgages, installments for the purchase of cars, appliances and other items, but also bills. For this reason it happens more frequently that banks turn to the courts to recover the money lent, but also that those who had requested a loan sue the banks, accusing them, for example, of applying usury rates, or of practicing compound interest. i.e. to calculate the interest also considering the overdue and unpaid interest, in addition to the initial capital.

Increasing in number, judgments affecting the actions of banks have become increasingly subject to interpretation. It happened that different courts took different decisions on the same legal issues. For more than twenty years, for example, jurists have been debating whether or not the default rates, i.e. those that a person pays when he or she is in arrears on payment, should be relevant in alleged cases of usury.

There are those who believe that default rates must also be counted to assess whether the interest requested by a bank falls within the limits of usury rates and those who argue the opposite, considering them particular rates that do not have to respect the threshold. Since the jurisprudence on the subject is complex, and often contradictory, the interpretation is often left to the individual judge who chooses, based on his own convictions, which criterion to adopt to make a decision.

For this reason, the Alma Iura financial mathematics study center for law and the consulting firm Mauden, an Italian company that develops management, analysis and data security software, have jointly designed an algorithm that allows you to calculate probability on the outcome of a judgment in the field of banking law. The software that does these calculations, still under development, is called Quaero Statistic & Law.

Quaero collects all the sentences issued by the Italian courts on the matter of interest and manages to understand them because Mauden has built a model of synthesis between the complex language of the sentences and that understood by the software. The jurists who worked on the project, with the collaboration of a language scientist, have developed a series of acronyms that have been associated with the different legal concepts, which the software archives, compares and processes. But it does not only do this, nor does it limit itself to returning cases and sentences with corresponding keywords, but encodes and classifies the data contained in them with the method of statistical inference and using artificial intelligence technology. In particular, Quaero is able to provide a probabilistic calculation on how a single judge could decide in a specific litigation, based on the average of his decisions in past judgments.

Basically, lawyers expert in banking law will be able to use Quaero to compare the results of the various judgments and have a series of additional data and correspondences: this will allow them, on a case-by-case basis, to choose the defensive strategy with the highest probability of success.

Quaero can also calculate how a judge may have changed his orientation based on any judgments of other courts, such as, for example, in the case of a sentence of the Supreme Court with joint sections, i.e. at the highest level of judgment. In other words, Quaero takes into account the possibility that a particular judge, after having expressed himself dozens of times in favor of those who presented the case in disputes on a specific issue, may change orientation after the Supreme Court, on that issue, has instead favored the reasons for the defense. Among the criteria with which Quaero will provide his probabilistic elaboration on the outcome of the trial, there will also be an evaluation of how much, in the past, and on other issues, the same judge has changed orientation in his judgments following a sentence of the Supreme Court.

Marco Rossi, lawyer of the Veronese firm Rossi Rossi & Partners and president of Alma Iura, explains that Quaero's goal “is not so much to be an oracle, but to provide glasses to a nearsighted person: it is a tool that serves to enhance the skills of the lawyer, certainly not to replace him ». Rossi adds: “There are things the machine does better, like calculating a huge mass of data that a human would take a lifetime to analyze. The only alternative to do the same would be to employ hundreds of people at exorbitant costs. Quaero catalogs and processes with a speed impossible for humans “.

Rossi explains that today “those who deal with the massive management of litigation, such as that on impaired loans, have tens if not hundreds of thousands of cases per year on all the courts, and therefore know what a judge thinks on a given issue, and how he decides, it is very important ». Especially to be able to answer the first question that a client asks the lawyer when he presents a dispute: how many chances do we have to win? «In our matter», Rossi continues, «if you don't know the judge who will deal with the case, it is almost impossible to give an ex ante answer. With Quaero we will be able to give at least one in terms of probability “.

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Most Popular

To Top