AI In Judiciary: Supreme Court Panel Unveils Rigid Draft Rules After ‘Judicial Misconduct’ Row

NEW DELHI — In a major bid to protect the sanctity of the courtroom from algorithmic bias, the Supreme Court AI committee has released a stringent set of draft regulations related to AI In Judiciary that completely ban the use of artificial intelligence to determine judicial outcomes.

The preliminary draft of the ‘Regulations for Use of Artificial Intelligence (AI) in Courts, 2026’, made public on Wednesday, sends an unequivocal message to the judiciary: technology must remain strictly subservient to human judgment.

The sweeping rules arrive on the heels of sharp criticism from the top court regarding a recent trial court’s over-reliance on AI, which resulted in the use of non-existent, AI-generated case laws. At the time, a Bench headed by Justice P.S. Narasimha warned that such lapses went far beyond simple errors, labeling them as outright “judicial misconduct.”

The Red Lines: What AI Cannot Do

The draft regulations establish strict boundaries, completely barring AI from core judicial decision-making. Under these rules, AI systems are explicitly prohibited from:

• Determining Sentences: AI-assisted sentencing is barred without mandatory, ironclad human oversight.
• Risk Scoring: The tech cannot be used to assess flight risks, predict recidivism (reoffending), evaluate bail eligibility, or judge the credibility of witnesses and parties.
• Opaque Systems: Courts are forbidden from using “black box” or unexplainable AI models where the underlying logic cannot be verified.
• Profiling & Surveillance: AI cannot profile witnesses or litigants, nor can it be used for the continuous monitoring of judges, lawyers, or litigants unless specifically authorized by law.

“Applications involving higher levels of risk to personal liberty… shall be subject to correspondingly heightened safeguards, including mandatory human-in-the-loop requirements.” — Draft Regulations, 2026

Where AI is Permitted

While drawing a hard line on judicial decision-making, the panel recognizes the efficiency AI can bring to back-office bottlenecks. The draft permits AI integration in routine, administrative functions:

Permitted AI Functions
Case management & automated scheduling
Preparation of daily cause lists
Real-time transcription of court proceedings
Translation of judgments across diverse languages

Governance, Data Privacy, and Bias Controls

The proposed framework mandates that any data processed by judicial AI must comply with the Digital Personal Data Protection Act, 2023. Furthermore, the committee heavily emphasized constitutional protections, asserting that AI must not introduce or amplify systemic biases related to race, religion, caste, sex, gender, disability, or economic status.

The panel also cautioned against widening the country’s digital divide, ensuring that AI-driven tools remain accessible to rural and economically disadvantaged communities.

A New Judicial Watchdog

To steer policy and set future standards, the draft recommends establishing a high-powered, full-time “apex body” at the Supreme Court level.

As proposed, this body will be heavily staffed by judicial and technical experts nominated by the Chief Justice of India (CJI), including two Supreme Court judges, four High Court representatives, an official from the Ministry of Electronics and Information Technology (MeitY), alongside dedicated cybersecurity and finance experts.

Public Consultation Timeline

The Supreme Court AI committee—chaired by Justice P.S. Narasimha and featuring Justices Sanjeev Sachdeva, Raja Vijayaraghavan V., Anoop Chitkara, and Suraj Govindaraj—has thrown the draft open to the public.

Stakeholders, legal professionals, and citizens have until June 20, 2026, to submit their comments and suggestions before the regulations are finalized into law.