The Supreme Court's draft AI regulations for courts reflect a cautious but necessary approach to technology in the justice system. Critically examine their significance for judicial independence and access to justice.

Q. The Supreme Court's draft AI regulations for courts reflect a cautious but necessary approach to technology in the justice system. Critically examine their significance for judicial independence and access to justice. (15 marks, 250-350 words)

Released on 3 June 2026 for public comment, the draft 'Regulations for Use of Artificial Intelligence (AI) in Courts, 2026' make AI "strictly subservient to human judgment" [1], echoing the CJI's assurance that AI will never overpower judicial decision-making [5]. Their caution is warranted, though not yet sufficient.

Safeguards for judicial independence - Non-delegable adjudication: barring AI from judgments, sentencing, bail and credibility assessment preserves the judge's application of mind as the sole source of decision [1]. - Ban on opaque/unexplainable systems protects the right to a reasoned order under natural justice, preventing dependence on vendor-controlled black boxes. - Prohibition on AI surveillance of judicial officers shields decisional autonomy from institutional monitoring. - Regulation 43 disclosure of AI-assisted filings responds to the Court's own censure of reliance on hallucinated, non-existent citations [2].

Gains for access to justice - Permitted uses — legal research, translation, transcription, summarisation, case listing and litigant chatbots — target pendency and language barriers in regional courts [1]. - Barring predictive profiling and risk scoring protects witnesses and undertrials, who form roughly three-fourths of India's prison population [3], from bias embedded in historical data. - A permanent Apex Body with SC/High Court judges, MeitY and technical experts promises uniform standards across all courts [1].

Critical limitations - The text remains a draft, without statutory backing or sanction for non-disclosure. - Uneven digital infrastructure across subordinate courts risks paper compliance rather than real audit. - "Assistive" AI still anchors judicial reasoning through summaries and drafts — an influence the draft does not address. - The Committee was reconstituted only in December 2025 [4]; auditing capacity and the interface with data-protection law are untested.

The draft rightly fixes AI as an aid to, not a substitute for, the judicial conscience while opening its efficiency gains to litigants. Notification with periodic algorithmic audits, judicial training and infrastructure parity across courts would convert this caution into credible protection of the Article 21 guarantee of a fair trial.

(~330 words)

Sources: 1. Supreme Court of India — Notices and Circulars (Draft Regulations for Use of AI in Courts, 2026) — draft's assistive-only principle, permitted and prohibited uses, Regulation 43, Apex Body composition 2. Supreme Court of India, Judgment dated 2 July 2026 (2026 INSC 668) — reliance on AI-generated fake/hallucinated case law by courts 3. NCRB, Prison Statistics India — share of undertrials in India's prison population 4. Supreme Court Reconstitutes AI Committee to Oversee Adoption and Deployment of AI Tools, Akashvani News (Dec 2025) — reconstitution of the SC AI Committee 5. Supreme Court Assures Judges Exercising "Utmost Caution" on Use of AI in Judicial Process, Akashvani News (Dec 2025) — CJI's assurance that AI will not overpower judicial decision-making