In a packed conference room at the Supreme Court complex, senior officials, researchers, and judges gathered yesterday for the release of a sweeping White Paper on Artificial Intelligence and Judiciary, a document that-quite frankly-felt less like a technical report and more like a roadmap for the next decade of Indian justice delivery.
The report, prepared by the Supreme Court’s Centre for Research and Planning (CRP), traces how courts worldwide have shifted from basic digitisation to far more advanced machine-learning tools. One official described it as “a moment where courts must pause, look around, and decide what technology is safe to bring into our halls of justice.”
Global Lessons… and Warnings
The paper opens with a tour through other nations’ AI frameworks—from UNESCO’s ethical guidelines to risk-based AI classification in the EU and Canada. Some countries use AI to summarise case files, others for tracking cross-border investigations or measuring sentencing trends. But buried between the success stories are clear red flags.
The paper bluntly notes several dangers: AI hallucinations producing fake case citations; algorithmic bias that can unfairly tilt outcomes; and deepfakes that may contaminate evidence streams. “Speed cannot come at the cost of justice,” the authors remind, emphasising that judges must remain the final authority at all times White Paper on AI.
India’s Own AI Journey
Where India stands today is surprisingly impressive. Tools like SUPACE (for fact extraction), SUVAS (for translation), and the Court’s AI transcription system (TERES) already exist in some form. The white paper carefully explains how these systems work, clarifying that technologies are used as aids, not as substitutes for judicial reasoning.
One interesting insight: machine-learning models may soon categorise incoming filings into clusters-criminal appeals, service matters, tax cases-making registry work faster. But the report is careful to warn that such automated sorting should always be checked by humans.
The Heart of the Document: Ethical Frameworks
A big chunk of the paper digs into ethics-exactly where Indian judges tend to worry the most.
The authors push six firm principles for any AI system:
- Human in the Loop
- Accuracy & Verification
- Confidentiality & Privacy
- Bias Prevention
- Clear Disclosure
- Restricted Roles for AI in core judicial functions White Paper on AI
One judge at the event told me off the record, “If a tool writes even a draft that I rely on, I must know where each sentence came from. Otherwise, it’s not law-it’s guesswork.”
Institutional Safeguards: Committees, Training, and In-House Models
The paper goes beyond theory, recommending concrete structural reforms:
- An AI Ethics Committee inside every court
- Development of in-house AI tools instead of depending on commercial models
- Mandatory training for judges, clerks, and registry staff
- Disclosure norms requiring users to state when AI has been used in preparation of filings
The report repeatedly stresses that free, public AI tools pose severe risks because they store user data, creating confidentiality nightmares-which, as the authors dryly note, “is unacceptable in judicial work of any kind.”
Court’s Decision / Order
The white paper concludes by formally recommending phased, ethics-anchored AI adoption, subject to strict safeguards, independent audits, curated datasets, and uncompromising human oversight at every stage of judicial decision-making. It emphasises that AI may support efficiency, but cannot-and must not-replace judicial discretion White Paper on AI.
Document Title: White Paper on Artificial Intelligence and Judiciary White Paper on AI
Issuing Authority: Supreme Court of India – Centre for Research and Planning (CRP)
Document Type: Policy / Research White Paper
Publication Date: November 2025