Introduction
Indian courts have faced the menace of delays and pendency for long. When a litigant walks into a court of law they do not merely expect a correct verdict. They also expect a prompt verdict ( Hussainara Khatoon v. Home Secretary State of Bihar AIR 1980 SC 81). Meanwhile the pendency of arrears remains a matter of great concern for the judges lawyers litigants and even the state. The National Judicial Data Grid currently stores around 5 crore pending civil and criminal cases across all the courts under its jurisdiction. Of these many cases have been pending for more than five years.
Meanwhile Artificial Intelligence (AI) emerges as a potential solution to many of these problems. Ai can help with document management translation transcription legal research case classification scheduling and even identification of old or urgent cases. The Supreme Court and e Courts have already started exploring some of these options. The Department of Justice has noted the use of Ai Machine Learning Optical Character Recognition and Natural Language Processing in e Courts applications for tasks like intelligent scheduling or filing information retrieval and communication with litigants.
On the other hand Ai is not a cure all for judicial delays or on hold. Courts are the ultimate fact finding organs which deal with liberty property reputation and fundamental rights. An incorrect algorithm or an erroneous AI generated authority therefore has far greater ramifications than just a technological glitch. The very recent case of Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. decided by the Supreme Court in July 2026 saw the intervention of AI in the working of tribunals and commissions leading to the generation of spurious and hallucinated case laws.
Therefore the debate must not be about whether AI should be used in judicial processes or not but rather about how AI could help speed up the justice delivery without undermining the rights of the citizens or bypassing the existing judicial or constitutional procedures.
Understanding Judicial Backlog
Judicial backlog and pendency of cases can be due to various reasons such as vacancies infrastructure issues numerous adjournments service of summons litigious culture complicated procedures shortage of personnel and delays in investigation or production of evidence.
The Supreme Court has often taken notice of this issue and acknowledged the fundamental right to a speedy trial in Hussainara Khatoon v. Home Secretary State of Bihar AIR 1980 SC 81. The Supreme Court also reiterated the same in Imtiyaz Ahmad v. State of U.P. (2012) 2 SCC 688. Therefore it is not merely a question of getting rid of the arrears but of dealing with several systemic issues. It is a matter of rights and as such it needs to be approached with caution.
At the same time it must be remembered that AI should not be deployed to simply get rid of pendency by obviating the need for judicial or quasi judicial processes. A system of courts is not designed to dispose of cases as quickly as possible but rather to deliver justice as per the existing procedures and protections.
How Can AI Help?
AI can offer solutions in a variety of areas. One of the prime advantages of AI is that it can handle repetitive tasks in a faster manner.
1. Legal Research and Document Scanning
Indian judiciary handles a huge volume of cases each of which generates numerous documents. AI can help extract relevant information from these documents for the purposes of reference research or record keeping. AI can help shorten the time taken in legal research substantially by quickly scanning through verdicts statutes and pleadings citing similar facts law or precedent.
The Supreme Court has already experimented with AI based legal research as well as other applications in the e Courts project. AI can also help in translation and related matters. The technology is already being explored for the purposes of judicial translations through the Supreme Court's SUVAS scheme.
However it must be noted that AI can serve as a research assistant but not as a decision maker. Therefore while legal research can be quickened by AI driven scanning and analysis the onus of applying the law and delivering the judgment rests with the judges.
2. Translation and Language
Indian judicature has to deal with multiple languages. Therefore it is natural that a lot of time is spent on translating documents. However translation can be a time consuming and even error prone process. The Supreme Court has developed the SUVAS system to help with translation and it has noted the usefulness of AI in this regard.
The Supreme Court also acknowledged the challenges and opportunities presented by technology to access justice in all languages. As such a technology driven approach to translation can be a major step towards making the judiciary more accessible. However it must be noted that AI needs to be regulated appropriately. After all the technology is only an aid and there must be checks to ensure that the accuracy of the final documents is not compromised. Such a system would be particularly helpful to litigants who are not fluent in Hindi or English.
3. Case Management
AI can be used for sorting out old cases managing the registry identifying missing documents and similar purposes. It can help identify senior citizens undertrials and other vulnerable groups needing special assistance
As per government documents regarding e Courts Phase III AI can be used to help manage and track cases more efficiently. It can help in intelligent scheduling prediction and management of old cases while focusing on vulnerable groups.
As such AI can help the registry system be more efficient and organized. Courts would then be able to switch from a reactive to a proactive mode.
4. Transcription
One of the most mundane tasks of the judiciary involves transcription. AI can help automate this process by converting speech to text thus reducing the workload of the court staff significantly. For instance the Supreme Court is already experimenting with AI driven transcribing of arguments and court proceedings.
While this application of AI might seem less exciting it could have a major impact on the efficiency of the judiciary. After all it helps to reduces the burden of the staff without requiring them to take on any decision making responsibilities. It is a win win for both the judges and the support staff.
The Legal and Constitutional Challenges Involved are:
Technology which cannot be deployed in Indian judiciary without due consideration of the legal and constitutional issues involved. Some of the issues include:
1. Algorithmic Discrimination
Ai systems are trained on data sets which can contain certain biases. As such it is entirely possible that Ai would inherit these biases and follow them while taking action or making recommendations. For instance an Ai based system for prioritizing cases might contain discriminatory features thus failing the test of Article 14.
It must be noted that the Indian Constitution prohibits the State from discriminating against any citizen on the basis of religion race caste gender or any similar grounds. Therefore it would be illegal for an Ai to discriminate on the same grounds simply because it was programmed to do so. However if an Ai system does indeed contain discriminatory features it would violate the basic structure doctrine.
2. Privacy and Confidentiality
Ai systems can be fed sensitive data by their developers. As such there is a risk that confidential or secret information might end up being shared publicly. Courts deal with a vast number of sensitive cases ranging from medical and financial issues to personal disputes criminal cases and child custody matters. As such there must be a strict policy of confidentiality when it comes to deploying such systems in the judiciary.
The Supreme Court has already recognized the fundamental right to privacy in Justice K.S. Puttaswamy (Retd.) v. Union of India (2017) 10 SCC 1. Therefore even the judiciary must respect the right to privacy when dealing with personal data. The Digital Personal Data Protection framework also recognizes this issue and addresses it in depth. The same principles would apply to Ai systems dealing with such data in the courts. After all the judiciary must ensure that the tools it uses do not compromise sensitive data.
3. Hallucinations and Bogus References
It is rather common for large language models to generate content with fake references. This means that an Ai model can provide a solution with a citation to a fabricated source or the wrong source. This poses serious risks for the judiciary especially in reference to legal authorities or cases.
The recent case of Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. decided in July 2026 provides rather strong grounds for concern. The case involved orders passed by the National Company Law Tribunal and the National Company Law Appellate Tribunal both of which had relied on six spurious non existent or wrongly cited statutes. The Supreme Court later overturned these orders noting the need for caution in reliance on Ai.
Therefore Ai driven legal research assistance can be extremely risky specifically when it comes to authorities. Every reference generated by Ai must be subjected to a reality check in order to ensure its authenticity. In the case of hallucinated references the issue might seem to be one of technological oversight rather than anything else. However it cannot be denied that a wrong authority can impact the decision making process of a tribunal or a court of law.
The Regulatory Environment in India
India does not have any law on Artificial Intelligence. There is a complex legal and policy framework that covers different aspects of artificial intelligence (AI) technologies. These laws and policies include the following:
-
The Constitution of India (1950)
-
Bharatiya Nagarik Suraksha Sanhita, 2023 (
-
The Digital Personal Data Protection Bill, 2023
-
Information Technology Act, 2000
The laws cover various aspects of the use of Ai and leave room for additional laws in the future. It should also be noted that a number of departments and agencies are already working on a variety of issues related to AI.
The Ministry of Electronics and Information Technology has recognized the risks posed by systems based on Ai, and has also emphasized the need for responsible innovation. MeitY has come out with recommendations on responsible Ai, including ways of addressing bias, promoting explainability, protecting privacy and conducting audits. Similarly Supreme Court has also been dealing with these issues and has sought suggestions on the regulations for use of Artificial Intelligence in courts 2026.
This is a step in the right direction as judiciary sector has its own unique requirements in terms of Ai use. Regulations on Ai in general may not be enough to address the concerns of the courts. Therefore, there must be a separate set of rules for judicial applications.
Ai must serve not replace the judiciary
The Judiciary in India should use Ai equipment as a tool to serve the cause of justice keeping in view the rights of all its stakeholders. The best approach would be a human in a loop approach where Ai would take care of simple tasks leaving the complex ones for human intervention. Note that the final authority to decide on matters of law should always be vested in a judicial officer or tribunal. This would ensure that the rights of the citizens or independence of the judiciary is not compromised. In fact, the recent judgment in the case of Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd delivered in July 2026 emphasizes the value of the human element in the use of AI in the judiciary.
It should be stressed that assistance and substitution are two entirely different concepts. So, a judge using Ai to find relevant precedents would still be acting as a decision maker, a totally different proposition altogether would be an Ai system directly delivering verdicts.
Way forward
The potential of Ai to be of great help to the Indian judiciary is enormous. It can be utilized for a number of purposes, such as research assistance, document scanning, transcription, and translation. It can also help to manage court schedules and some other related issues. At the same time it is essential that proper safeguards are put in place to guaranty the responsible use of Ai. Some of the proposed measures are as follows:
-
Create a comprehensive regulatory framework for judiciary on all aspects of Ai and on different types of applications of Ai .
-
Try to use Ai for more "routine" tasks, at least initially. There's no burning need to use AI in core decision-making activities.
-
Ensure all references generated by Ai are properly authenticated. In fact, this point has been reiterated by the recent judgment of the Supreme Court in Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. 2026 INSC 668.
-
Set up proper audit mechanisms to check if Ai platforms are free from bias or any other issues. The source code may not have to be made public , but the auditing process should be thorough enough to be able to catch potential problems .
-
Pay proper attention to privacy and confidentiality concerns. Do not share sensitive data with unauthorized parties.
-
Ai must be viewed as a part of the larger judiciary reforms and not as a replacement to the reforms. Technology itself cannot solve the problem of judicial backlog. It can only complement the efforts of reforming procedure increasing infrastructure and improving support system. The vacancies, the judicial appointments and even the general administrative issues of the courts have to be dealt with separately.
Conclusion
Ai presents India with an unprecedented opportunity. From helping with legal research to managing schedules translation and transcription Ai has the potential to reduce the burden on the Indian judiciary considerably. However technology driven solutions cannot undermine the fundamental rights of the citizens. The judiciary exists to protect these rights and should strive to deliver justice to every individual without fail.
As such any technology deployed in the judiciary must adhere to the existing rights framework especially the right to life and liberty. At the same time the judiciary must recognize the potential of technologies like Ai in addressing contemporary issues. The recent case of Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. decided by the Supreme Court in July 2026 actually demonstrates that Ai can be a useful tool in supporting judicial work. However the same case also underlines the potential risks and dangers of adopting such systems without appropriate safeguards.
References
Legislation
- Constitution of India 1950
- Bharatiya Nagarik Suraksha Sanhita 2023
- Digital Personal Data Protection Act 2023
- Information Technology Act 2000
Cases
- Hussainara Khatoon v. Home Secretary State of Bihar AIR 1980 SC 81
- Imtiyaz Ahmad v. State of Uttar Pradesh (2012) 2 SCC 688
- Justice K.S. Puttaswamy (Retd.) v. Union of India (2017) 10 SCC 1
- Anita Kushwaha v. Pushap Sudan (2016) 8 SCC 509
- Salem Advocate Bar Association T.N. v. Union of India (2005) 6 SCC 344
- Pooja Ramesh Singh v. Jammu and Kashmir Bank Ltd. 2026 INSC 668
- Government Reports and Official Materials
- Supreme Court of India White Paper on Artificial Intelligence and Judiciary November 2025
- Supreme Court of India Unclogging the Docket Revised Edition November 2025
- Department of Justice Ministry of Law and Justice Government of India materials concerning AI and the e Courts Project 2025–2026
- National Judicial Data Grid pending case statistics
- Ministry of Electronics and Information Technology Annual Report 2024–25
- Official Government Sources
- Supreme Court of India Reports and Publications
- Department of Justice Ministry of Law and Justice Government of India
- India Code Government of India
- National Judicial Data Grid e Courts
- Ministry of Electronics and Information Technology Government of India.
- Academic and Research Sources
- PRS Legislative Research Impact of Emergence of Ai and Related Issues
- Various academic resources on Artificial Intelligence Judicial Administration Algorithmic Bias Privacy and Right to Justice