AI in courts as a Way Forward in Judicial Analytics

Introduction :

Imagine a scenario where a judge is armed with advanced analytics and artificial intelligence tools to handle each case on his/her docket. What if there is a method by which judges could be made aware of a potential delay in receiving police evidence in a murder case or the probability of a witness turning hostile on certain grounds? Vital information like this will help judges strategize and dispose cases faster. These scenarios are far from hypothetical – several countries are already using such technology in the judicial process. In order to carry out such predictions, a significant amount of data from various sources is required, stored in a centralized data warehouse that can communicate with each other.

Artificial Intelligence In Judicial Analytics :

In the context of the Indian judicial system, much has been said about judicial delays, pendency, and many ways to tackle these issues have been proposed. These include measures such as increasing judge strength, cutting down vacation time, and non-acceptance of frivolous cases. For a judiciary which has more than 3 crore pending cases, much more needs to be done.

There have been very few measures taken to assist judges in managing cases. Thousands of cases that are fairly similar in nature are being filed each year. There are very few attempts made to analyse how these cases are progressing at every stage. Judges need to be equipped with enough information about the possible paths that the case could take to help in faster disposal of the case. As mentioned above, data science and Artificial Intelligence (AI) are capable of assisting judges by predicting vital information regarding an ongoing case based on past cases of a similar nature. Several case details such as number of accused in  a case, date of filing of the charge sheet, number of witnesses examined during evidence stage, witness turning hostile, reasons for adjournments, First Information Report (FIR) details, quantum of punishment, compensation granted etc. which are recorded in the daily orders and the final judgment can be used for this analysis. Analysis of these variables can help judges make better strategic decisions which can help reducing delays in a case.

AI can be particularly useful at the evidence stage. Research suggests that evidence stage is vital in the life cycle of a case and takes up a large proportion of the court’s time.  Multiple adjournments are sought during the evidence stage for various reasons such as extra time sought by the advocates/parties, delay due to the Investigation Officer, absence of witnesses etc. Artificial intelligence tools would be helpful in predicting potential delays as judges would know before-hand the major causes for delay for a particular type of cases at a given stage and the judges will be able to list matters and handle their caseload accordingly. It can help judges predict factors that will delay the trial. For example, if the probability of witnesses turning hostile is higher in a certain type of case, the judge can proactively take steps to prevent witnesses from turning hostile, such as ordering for police protection.

However, despite the intelligence of such algorithms, there will be inaccuracies as is the case with all unstructured data sources. In order to build a robust algorithm, data needs to be structured. Unfortunately, most of the relevant data on the e-courts website (official website containing case related data from across India) is not in a structured format. Data which is in heavy text format e.g. order sheets & judgements needs to be recorded in a more structured manner.

Conclusion :

The e-Courts portal has been one of the most significant steps taken for judicial reforms in the recent era. In the decade since its launch, it has made a vast trove of data available online, at no cost for users. Unfortunately, as mentioned earlier, a lot of important data points in the daily orders and judgments are in a text heavy format and are difficult to analyse. Despite the ordeal of working with unstructured data, we should start exploring ways to use it for judicial analytics. Although Artificial Intelligence could help us understand the present data, accuracy of such analyses can only be higher if we bring some change to the way this data is being recorded. Once the process of making this data available on e-Courts has been changed, it can be used for a variety of analyses that will be useful for judges and court managers. This will herald a new era in the field of judicial analytics in India.

The views expressed in this article are solely those of the author’s and they do not represent the views of DAKSH.