On 3rd June (Wednesday), a draft set of rules governing the use of artificial intelligence (AI) in Indian courts has been released by the Supreme Court. It is titled, “Regulations for Use of Artificial Intelligence (AI) in Courts, 2026.” Furthermore, stakeholders and general public are asked to provide feedback on the same by 20th June.
The draft was created under the auspices of the Supreme Court’s AI Committee. It aims to create a framework for the judiciary’s responsible use of AI based on the values of human primacy and judicial independence, rule of law, fairness and non-discrimination, transparency and explainability, accountability, auditability and continuous oversight, data protection and privacy, purpose limitation, proportionality, inclusivity and accessibility, data integrity, cyber security, presumption in favour of responsible AI adoption along with innovation over restraint.
The court stated, “They shall apply to the use, deployment, or integration of Artificial Intelligence in any judicial, adjudicatory or administrative function of the Supreme Court of India, High Courts, and all Courts including the Tribunals and statutory Commissions performing adjudicatory functions, within the territory of India.”
A crucial part of the proposed restrictions is the premise that AI systems remain firmly subject to human judgement. “Every AI System shall function solely in an assistive capacity and shall not supplant or compromise the independent exercise of judicial authority by a duly appointed judicial officer. The ultimate authority to determine matters of law, fact and justice shall vest exclusively in the judicial officers of the competent jurisdiction,” the draft read.
It further highlighted, “Accountability for all decisions made by any officer with the assistance of AI shall rest exclusively upon such officer and it shall not be permissible to invoke the outputs of an AI System, the opaqueness of a Black Box system, or the occurrence of hallucination, as a ground for avoiding accountability for a palpably incorrect, illegal, or harmful decision.”
The draft stated that courts should prioritise deploying AI systems and tools to improve access to justice, minimise delays, and boost administrative efficiency. Responsible usage of AI in court procedures is expected unless proven otherwise.
A variety of tasks, such as case management, scheduling, cause-list preparation, transcription of proceedings, translation of legal documents, legal research, citation verification, document summarisation, accessibility services for people with disabilities, anonymisation of judgements and court records, and administrative tasks, are permitted by the regulations. However, prior authorisation and human oversight would be necessary for the majority of these purposes.
Throughout its use, every AI system used in court procedures must be continuously monitored and subject to recurring technical, legal, and ethical audits. Sufficient procedures must also be put in place to identify, record, and correct mistakes, malfunctions, and prejudices.
AI can be utilised for docket prioritisation, hearing scheduling, cause list development, and case administration. Moreover, it can be used for legal research, citations, administrative tasks, automated transcription of court proceedings, and translation of judgements, orders, and pleadings (subject to human verification).
The restrictions of AI usage and an Apex Body to prevent violations
AI systems will not be applied to decide cases or make court decisions on their own. It is specifically prohibited to employ AI for risk scoring, which includes forecasting recidivism, evaluating flight risk, determining bail eligibility, and rating witness reliability. Additionally, the restrictions forbid the use of opaque and inexplicable AI systems in situations pertaining to rights or personal liberty, monitoring judicial officers or litigants, or forecasting future behaviour of parties or witnesses.
“No judicial outcome (including any judgment, order, or finding of fact or law) shall be reached through Algorithmic Decision-Making alone or solely on the basis of AI-generated information, data, or analysis and the human judicial authority shall be the determinative authority in all adjudicative decisions,” the draft outlined.
It further conveyed, “No AI System shall be used in any manner that may compromise the confidentiality of judicial deliberations or the independence of the judicial decision-making process.” Any infraction of a prohibition must be reported immediately to the AI Secretariat and brought before the AI Committee. Following a thorough investigation, the AI Committee will order any necessary corrective action, including suspending the relevant AI system.
“These prohibitions are absolute and non-derogable, and shall not be subject to relaxation or modification by any authority under these regulations,” the draft directed. To oversee and promote innovation, integration, governance, supervision, standard-setting, and policy development on artificial intelligence in the court, the Supreme Court will establish a permanent, full-time Apex Body.
It will be composed of two Supreme Court judges, two Chief Justices of the High Court, two High Court judges, one member from a nationally significant institution, an officer in the Ministry of Electronics and Information Technology not lower than the rank of Joint Secretary, an expert in finance, an expert in cybersecurity, one or more standing advocates with knowledge of technology-related laws, and an AI professor from National Judicial Academy Bhopal.
Adherence to Digital Personal Data Protection Act, 2023
The proposed system places stringent controls on judicial data and requires adherence to the Digital Personal Data Protection Act, 2023. In addition to being subject to contractual duties regarding data ownership, cybersecurity, audit rights, and responsibility, private technology providers wishing to take part in court AI systems would need prior approval.
“All proposals for engagement of private entities in connection with AI Systems shall, prior to approval by the Appropriate Authority be subject to a comprehensive evaluation covering technical capability, legal compliance, ethical standards, data security practices and financial standing.,” the draft mentioned.
Courts would maintain ownership or a perpetual royalty-free license over such systems and their outputs in cases when AI tools are created utilising court data or resources.
AI Committees in the Supreme Court and each High Court, specialised AI Secretariats, regular technical, legal, and ethical audits of AI systems, incident reporting systems, cybersecurity protections, and specialised training programs for judges, attorneys, and court employees are all included in the framework. Importantly, the draft clarified that attorneys or litigants cannot avoid accountability by attributing false or deceptive filings to artificial intelligence.
The rules mandate that lawyers, judges, and court employees participate in frequent training sessions on the ethical, legal, and technological implications of AI. AI bias, hallucinations, data security, cybersecurity, and reporting procedures for AI-related events would all need to be covered in training sessions. The framework mandates that courts set up emergency and backup plans to guarantee business continuity in the event that an AI system malfunctions, fails, or is unavailable.
The draft firmly pronounced, “Nothing in these regulations shall preclude any person from seeking such legal remedies as may be available under any law for the time being in force before a competent Court in respect of any harm caused by the use of AI in Court processes.”
Kerala High Court’s AI guidelines
Kerala High Court also developed an AI usage guideline last year that expressly forbade the district judges from using such tools for decision-making or legal reasoning. In light of the growing accessibility and availability of such software tools, it issued the “Policy Regarding Use of Artificial Intelligence Tools in District Judiciary” for the responsible and limited use of AI in judicial tasks of the state’s district judiciary.
The policy mandated that the use of AI tools for official purposes should stick to the standards of accountability, transparency, fairness and confidentiality protection. It recommended for avoiding the use of cloud-based services, with the exception of authorised AI tools, careful verification of the outcomes produced by such software, including translations, and constant human oversight of their use.
“AI tools shall not be used to arrive at any findings, reliefs, order or judgement under any circumstances, as the responsibility for the content and integrity of the judicial order, judgement or any part thereof lies fully with the judges,” it asserted.
The phenomenon of AI “hallucination” in court orders
An Advocate Commissioner was assigned by another junior civil judge in Vijayawada to survey a disputed land and ascertain whether it was within the parameters of a particular sale document during a case hearing. The commissioner’s report was challenged by the defendants, who objected to its conclusions. Citing four rulings from the Supreme Court, the court rejected these objections in August 2025.
When the order was contested before the Andhra Pradesh High Court in January, it became clear that none of these rulings actually existed. The additional junior civil judge acknowledged in a report to court that this was the first time she had utilised an AI tool for case law research. The matter reached the apex court which noted that a verdict stemming from fictitious rulings “is not an error in decision making but misconduct” which entails “legal consequence.”
There have also been instances in the past where authorities and litigants have relied on “AI generated non-existing, fake or synthetic alleged judgements” a termed by the Supreme Court.
A special leave petition was dismissed by the Supreme Court on 13th February, after it was discovered that the petitioner had cited rulings which never took place. A petitioner was fined Rs 50,000 by a Bombay High Court bench in January 2026 for mentioning a similarly bogus case in written submissions.
A different Bombay High Court bench overturned an Income Tax Department assessment order in October 2025, ruling that it had spiked a company’s revenue by more than Rs 22 crore based on three “completely non-existent” court verdicts. A Delhi High Court petitioner withdrew their plea in September 2025 when opposing attorney noted that “some of the judicial precedents cited on behalf of petitioner do not even exist, and, in some of the precedents, the quoted portions do not exist.”
