Translation to paperwork: How AI is making its case in courts, tribunals & errors, Pg9

Key Highlights:

• TERES, a real-time transcription system, is used in the Supreme Court for Constitution Bench hearings.
• Adalat AI is operating in High Courts of nine states and has pilot projects in five others to streamline case management.
• The Supreme Court uses SUVAS to translate judgments into 18 vernacular languages; over 83,783 judgments translated as of March 2025.
• AI tools are being adopted by lawyers and law firms for drafting, document review, and legal research.

Detailed Insights:

• TERES improves accuracy in recording court proceedings by converting oral arguments into text with a two-layer system involving AI and human review.
• Adalat AI assists judges by consolidating cases from different stages into a single window, reducing time spent on locating files and understanding context.
• The Kerala High Court uses AI to screen vakalatnamas, pleadings, and annexures at the filing stage, processing around 46,000 cases to reduce errors.
• SUVAS aids in translating judgments and orders into regional languages, enhancing accessibility of legal information.
• AI adoption in law firms is increasing for drafting legal documents, reviewing large volumes of text, and synthesizing legal precedents.
• Security measures for AI systems in courts include deployment within sovereign borders, in-house AI model development, and explicit consent for data access.
• Infrastructure constraints like unreliable connectivity and power cuts are considered in the development of AI tools for courts.

Key Concepts Involved:

• Artificial Intelligence (AI): Simulation of human intelligence processes by computer systems.
• Transcription: Converting spoken language into written text.
• E-filing: Submitting court documents electronically.
• Vakalatnama: A document authorizing a lawyer to represent a client in court.