Lawyers can use AI if...Supreme Court releases draft regulations for use of AI in courts, invites inputs

The Supreme Court has released draft regulations governing the use of Artificial Intelligence (AI) tools across courts in India.

As per the same, the Court has proposed that lawyers may use AI tools to prepare pleadings and evidence but they must tell the court when they do so.

The draft Regulations for Use of Artificial Intelligence in Courts, 2026 permit AI-assisted legal research, drafting, translation, transcription and case management, but make it clear that judges alone will retain the authority to determine questions of law, facts and justice.

Regulation 43(3) of the draft regulations states that where an AI tool is used by any party or legal representative in the preparation or submission of any document, pleading or evidence, the AI-assisted character of such material must be disclosed to the court at the time of submission.

Thus, the framework encourages the adoption of the technology while drawing clear limits around its use in the justice delivery system.

Regulation 43(4) also empowers courts to seek details of the AI system used, the nature and extent of AI assistance taken and the steps adopted to verify the accuracy of any AI-generated content.

The draft makes it clear that lawyers or litigants cannot escape responsibility by blaming AI for false or misleading filings.

Regulation 43(6) states that if any document, pleading or evidence submitted to a court is found to be fabricated, false, misleading or inaccurate because of its AI-generated character, the person submitting it will bear full responsibility.

Such a person cannot rely on the AI-generated nature of the output as a defence. The court may also pass such orders as it deems fit against the responsible person.

The proposed regulations would apply to the Supreme Court, High Courts, subordinate courts, tribunals and statutory commissions performing adjudicatory functions across the country.

It has been published by the Supreme Court’s AI Committee and has been placed in the public domain for comments and suggestions till June 20.

The draft regulations set out in detail the functions for which AI may be deployed in courts, as well as a list of activities for which its use would be completely prohibited.

What AI can do

The draft regulations expressly permit AI to be used for a range of judicial and administrative functions, subject to approval and human supervision.

Among other things, AI may be used for:

• Legal research and precedent retrieval;

• Citation verification;

• Summarising pleadings, judgments and documents;

• Translation of judgments, orders and legal documents;

• Automated transcription of court proceedings;

• Drafting assistance and generation of prescribed formats;

• Preparation of notices, summons and administrative documents;

• Cause list preparation and hearing scheduling;

• Case management and defect scrutiny;

• Record management;

• Judicial administration and resource allocation;

• Accessibility tools for persons with disabilities; and

• AI-powered assistants and chatbots to help litigants access court services.

The regulations also state that courts should actively explore AI tools that improve access to justice, reduce delays and enhance efficiency.

What AI cannot do

The framework places several absolute restrictions on the use of AI in courts.

AI systems cannot:

• Decide cases or judicial outcomes;

• Pass sentences;

• Determine bail eligibility;

• Assess flight risk or likelihood of reoffending;

• Evaluate witness credibility;

• Predict future conduct of litigants, witnesses, lawyers or accused persons;

• Conduct surveillance of judges, advocates or litigants;

• Interfere with judicial deliberations; or

• Use opaque “black-box” decision-making systems in matters affecting rights or personal liberty.

The regulations specifically prohibit courts from using AI-based risk-scoring systems to assess whether an accused may commit future offences, flee from justice or qualify for bail.

Disclosure of AI-assisted filings

The draft regulations introduce disclosure requirements for parties using AI tools in litigation.

Where AI is used in the preparation of pleadings, documents, submissions or evidence, the AI-assisted nature of the material would have to be disclosed to the court.

Courts may also require disclosure of the AI system used, the extent of AI assistance provided and the steps taken to verify the accuracy of the AI-generated content.

The regulations further provide that AI-generated material cannot be produced before a court as an independent source of evidence without full disclosure of its AI-generated character.

Any person using synthetic or AI-generated information in judicial proceedings would also be required to disclose such use.

Human judges remain responsible

The regulations are built around the principle of “human primacy”.

They state that AI systems must function solely in an assistive capacity and cannot replace or compromise the independent exercise of judicial authority.

“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 draft states.

The regulations further provide that the ultimate authority to determine questions of law, fact and justice must remain exclusively with judges.

Even where AI tools are used, accountability for decisions would continue to rest with the judicial officer or court official concerned.

The draft specifically states that officers cannot rely on AI outputs, the opacity of a system or AI hallucinations as a defence for an incorrect, illegal or harmful decision.

Recognition of AI hallucinations

The draft regulations expressly recognise the problem of AI “hallucinations”, describing them as instances where AI systems generate plausible but inaccurate, fabricated or misleading information, including incorrect legal precedents, statutory provisions or facts.

AI-generated outputs would ordinarily have to be verified before being used and would remain advisory in nature.

New governance structure

The regulations propose the creation of a permanent apex body at the Supreme Court level to regulate and oversee the use of AI in the judiciary.

The proposed body would comprise Supreme Court judges, Chief Justices and judges of High Courts, technology experts, cybersecurity specialists, finance experts and advocates with expertise in technology-related laws.

It would be tasked with setting national standards, approving AI systems, coordinating with High Courts and publishing annual governance reports on AI adoption in courts.

The framework also proposes AI Committees and dedicated AI Secretariats in the Supreme Court and every High Court to approve AI tools, monitor compliance, investigate incidents and oversee implementation.

A Centre of Research and Excellence on Artificial Intelligence (CoRE-AI) is also proposed to conduct research, evaluate AI tools and provide technical support to the judiciary.

AI audits, registers and incident reporting

Every AI system deployed in courts would have to undergo technical, legal and ethical audits at intervals not exceeding one year.

Courts would also be required to maintain an AI Register recording approved AI tools, their purpose, audit findings, conditions attached to their use and any AI-related incidents.

The regulations additionally propose an AI Incident Database to record malfunctions, errors, bias, security breaches and other incidents involving AI systems.

Any malfunction, error or bias with potential legal consequences would have to be reported immediately to the AI Secretariat.

The framework also requires annual transparency reports from High Courts, tribunals and commissions detailing AI systems in use, audit outcomes, incidents and compliance measures.

AI content verification authority

The draft regulations propose the creation of an AI Content Verification Authority to oversee standards, tools and verification protocols for generative AI content used in court processes.

Restrictions on private AI vendors

The framework places safeguards on the engagement of private technology companies.

No private entity would be permitted to participate in AI systems used by courts without prior approval from the competent authority.

Contracts with vendors would have to include provisions relating to ownership of court data, restrictions on the use of judicial data, audit rights, cybersecurity safeguards, disclosure obligations and liability for AI-related harm.

The regulations also provide that where AI tools are developed using court data or court resources, courts must retain ownership of, or a perpetual royalty-free licence to, such tools and their outputs.

Data protection and cybersecurity

The draft regulations require compliance with the Digital Personal Data Protection Act, 2023 and other applicable laws.

Personal data cannot be used to train, test or refine AI systems without prior approval from the competent authority.

Sensitive judicial data would be subject to heightened protections and could not be transferred to external systems without express authorisation.

The framework also mandates anonymisation wherever feasible, application of data minimisation principles and periodic cybersecurity audits of AI systems.

Training and emergency protocols

The regulations require regular training programmes for judges, lawyers and court staff on the technical, legal and ethical aspects of AI.

Training modules would be required to cover AI bias, hallucinations, data protection, cybersecurity and reporting mechanisms for AI-related incidents.

The framework also requires courts to establish emergency and fallback protocols to ensure continuity of operations if an AI system fails, malfunctions or becomes unavailable.

Presumption in favour of AI

Pertinently, the draft regulations adopt a policy of encouraging the responsible use of AI in courts rather than restricting it by default.

The regulations create a “presumption in favour of responsible AI adoption” and state that courts should actively explore AI tools that can improve access to justice, reduce delays and enhance administrative efficiency.

They also provide that any decision to restrict or refuse the use of an AI system should ordinarily be supported by reasons recorded in writing.

The draft further states that AI should be pursued as a catalyst for innovation in the justice delivery system and that, where appropriate safeguards are in place, an approach favouring responsible adoption should be preferred over restraint.

The Supreme Court has invited comments and suggestions on the draft regulations before finalising the framework.

Read the draft here.