It was revealed in 2016 that a United States (“US”) based law firm, Baker Hostetler, employed ROSS, the world’s first-ever robotic lawyer in the world. ROSS, a legal research robot built around IBM’s Watson cognitive computer, can collect data, make conclusions, and provide responses to a wide range of queries posed by attorneys. Chief Justice D.Y. Chandrachud stated that “Technology is here to stay for the future, forever” in response to a petition asking for the right to virtual hearings to be recognized as a fundamental right. It is evident that the artificial intelligence (“AI”) genie is already out of the bottle and that the presence of the same can already be seen in the sphere of law as well. Hence, by analysing the benefits and disadvantages of the same, it’s time to map out the potential future of the justice system through the AI-vengers Initiative.
When it pertains to AI, predictive analytics is a segment of machine learning that uses analytical algorithms and a data mining approach to examine prior data in search of patterns and associations that may be used to forecast future events. In the legal field, predictive analytics may be used to rummage through an abundance of data in search of patterns and insights that might guide attorneys and judges toward better outcomes.
Predictive Analytics: A Comparative Discourse
In order to analyse the potential of predictive analysis in the legal system, a comparative narrative of the same including major nations could be beneficial. The United States has embarked on numerous programs that use AI for a multitude of objectives with the intent of bolstering the administration of justice. Some U.S. courts utilize artificial intelligence in addition to traditional research techniques to ensure that their rulings are compatible and consistent. Case outcomes may be predicted by AI systems that scrutinize the data, operating a thorough analysis of the various statistics, including but not limited to criminal track record, social and economic standing, mental health, and physical health, etc. of the suspect, the US-based AI tool named Correctional Offender Management Profiling for Alternative Solutions, (“COMPAS”) provides assistance to the judges to better adjudge the cases before them.
Similarly in the United Kingdom (“UK”), with the intent of reducing the accumulation of cases in the crown’s court, the UK Ministry of Justice introduced the AI-based digital case system which glared through the fumes and proved to be a competent tool in the implementation of justice. It is also fundamental to consider China’s manoeuvre of AI in their justice system as an example, as they have been implementing it since the 1990s, a very recent example of the same court is the “Smart Court System” which conjures up with every desktop in the courts and assists the judges to deliver a preferable case analysis. In Columbia, for the first time in the history of the legal system, a judge used the AI tool ChatGPT to deliver a judgment, wherein he referred to the tool to get an analysis on the question of “Whether an autistic minor can be exonerated from paying fees for their therapies?”
This comparative analysis does provide us with the idea that AI in the justice delivery system has definitely attained world attention and every major nation has already been well indulged in the same. This transformation is a testament that signifies a paradigm shift in the manner in which legal procedures are conceived, executed, and adjudicated. The global entry into the justice domain makes it even more essential to analyse, what effect the same has had on the Indian legal system.
The Indian Perspective on Predictive Analytics
India did not lag behind when it came to garnering the use of AI tech in the justice delivery system. Recently, the Punjab Haryana High Court made history by using ChatGPT to convey a decision in a criminal case. During a bail hearing presided over by Justice Anoop Chitkara, the AI tool was questioned about the jurisprudence on bail in cases where the assailants have assaulted the victims with cruelty. ChatGPT responded that the decision on bail in such cases would depend on the specific circumstances of the case and the laws and regulations of the jurisdiction. This marks the first time an Indian court has used an AI-powered assistant to aid in a legal decision.
However, the predictive analytics tools are still considerably a road less travelled for India’s justice system. However, there have been some noteworthy pilot projects and initiatives being taken by different states to facilitate the justice delivery system with the aid provided by AI’s predictive analytics technique. An example of this is the AI tool used by the Maharashtra Police to detect and analyse crime data, predict potential crime hotspots, and deploy resources to prevent any probable crime. The program uses advanced algorithms to ensure accuracy and effectiveness in crime prevention.
AI Tools developed by the legal technology sector can gather and analyse massive amounts of data to make accurate case predictions. One such software, called “Premonition” purports to be able to forecast a lawyer’s chances of winning, the length of their case, and the judge they will be assigned to. Similarly, another software “LexMachina” provides access to a judge’s earlier rulings, providing insight on whether or not the case has a good chance of success. Similar results may be achieved with other AI tools like “Ravel Law”, “Intraspexion”, etc.
One of the potential benefits of predictive analytics and AI is improved accuracy in assessing risks, which can enhance public safety. The National Crime Records Bureau (“NCRB”) in India collects and analyzes crime data using the Crime and Criminal Tracking Network and Systems (“CCTNS”) system. This information is utilized to monitor the movements of known criminals and repeat offenders who pose a considerable threat to society. Haryana has been recognized for successfully implementing this system and ranks at the top.
The Odisha state police force has established a machine learning algorithm-based AI system, named “Crime Criminal Analytics and Prediction System” (“CAPS”) to determine the probability of an accused individual leaving the state. This approach has made the judicial and law enforcement decision-making processes more objective by reducing the role of subjectivity. Similarly, the state of Telangana has also adopted an AI-based predictive analytics system called the ICJS, i.e., Integrated Criminal Justice System (“ICJS”) which on analysing the available data and other facets shall create a profile of the suspect, and can predict the probabilities surrounding the re-commission of crime by the offender.
The Government Participation
There have been ongoing debates around the pros and cons of AI all around the globe, still a lot many nations have been proactive in inducing AI regulation policies for better induction of AI-based models in the governance mechanism. However, the position in India is that there hasn’t been any official AI policy being introduced by the Indian Government so far. In 2020, India released a draft of the “National Strategy for Artificial Intelligence” and this was a big step towards establishing the fact that the future does hold a strong AI policy for India.
In order to formulate an effective AI policy in India, the best way is to prepare a comparative analysis of the AI policies of other nations. The United States established a National Artificial Intelligence Initiative Office in 2021 wherein the primary focus shall be to multiply the efforts by America to establish its leadership in new-age tech such as AI. This office shall act as the centre for research in order to better regulate AI tech and integrate the same into the working of the federal government. India can also take inspiration from the Canadian Automated Decision Making Directive and the Algorithmic Accountability Act of the US in order to guarantee that AI is utilized in an equitable manner and to further avoid any algorithmic bias. Japan’s AI Utilization Promotion Act can also be a cue for India’s AI policy development in order to create a supportive environment for AI development through education.
One potential conflict of India’s prospected AI policy is with the recent Data Protection Bill introduced in India. Although the bill doesn’t categorically oversee the use of AI, several of its clauses are pertinent to AI and directly oppose how it facilitates the collection of personal data. The bill’s data protection standards and the full use of AI’s capability are clearly at odds with each other when viewed through the lens of AI. The bill lays down the exception of “public interest” and other “fair and reasonable” purposes wherein the government won’t require individual consent while utilising data. This exception, if interpreted carefully, could be useful to fill the gap in the use of AI for public purposes such as predictive analysis in the justice system. Data utilised to take input from an AI-based predictive analysis model could come under public purpose if the same is used for crime spot detection, bail, or other associated legal technicalities. Further, Section 18(2)(b) of the bill allows the government to utilise data as long as no decision-specific user profiling is done with the personal data.
The Way Ahead for Predictive Analysis in Law
The use of AI in the legal field is rapidly increasing and has gained significant attention, with UNESCO offering a course on ‘AI and the Rule of Law’ to educate legal professionals and other interested parties on the advantages and disadvantages of AI in the legal sphere. The importance that AI holds for the legal sphere across the world was also highlighted by Hon’ble Justice Huma Kohli when she stated that “AI is a game changer in the legal field and entails the potential to revolutionise the way the system works”, and hence laid down the point that AI shouldn’t be viewed as a threat, but rather as an opportunity.
As deliberated above, Indian states are already acting at the forefront in adopting AI-based predictive analytics in order to predict potential crime hotspots, risk assessment, determine appropriate sentencing, detention-related queries, and even biometric identification. For instance, the Patna High Court faced challenges in quickly allocating cases. To address this, MCIL developed a system that automatically assigns cases based on an algorithm that considers the type of cases each bench had previously handled and how efficiently they were resolved.
It is important to recognize that if biased information is fed into predictive analytics AI, it may perpetuate inequality and prejudice. This is a potential concern that should not be ignored. The criminal justice system may be detrimentally impacted if, for instance, an algorithm is trained on data that disproportionately targets specific populations or classes. Consequently, it is crucial to guarantee that new technologies are not applied in a manner that violates the due process, rights of people or is biased against any group.
Predictive Analysis could reap real fruits if applied effectively in the justice delivery system, for it has a great potential to assess and evaluate the risk posed by individuals awaiting trial, and hence assist the judiciary in deciding matters regarding bail and release more efficiently. The inclusion of AI interface into the justice delivery system in no way undermines the traditional techniques, rather it complements the traditional techniques and imbues the legal system with continuing durability. The comparative analysis provided above clearly lays down the picture as to how predictive analysis has already attained a durable global standing, and the reason for the same is that it acts as a catalyst for legal innovation.
Therefore, a well-rounded and extensive strategy is crucial to pledge that the application of AI and predictive analytics in criminal justice is uncapped, accountable, and in compliance with fundamental ethical standards. The implications on society as a whole, especially on vulnerable populations, must be assessed, and any inherent prejudices or unexpected repercussions must be addressed.