AI Could Clear 60% of Pendency If Routine Cases Move to Automation: Justice Manmohan
AI can take a meaningful bite out of India's case backlog-over 60 per cent-if routine, small-ticket litigation shifts to automated systems. Speaking at the India Law AI and Tech Summit 2025 in Delhi, Supreme Court judge Justice Manmohan argued that technology should absorb clerical load so judges can focus on complex legal questions.
He pointed to the obvious candidates: traffic challans, cheque-bounce matters, and other repetitive disputes that rarely require deep reasoning. Move these to AI-assisted platforms and the courts can concentrate resources on the core 40 per cent.
What Can Move First
- Traffic challans and standard compoundable offenses.
- Cheque-bounce matters where facts and statutory templates are largely uniform.
- Other form-based, rule-driven disputes with settled law and minimal discretion.
Justice Manmohan also highlighted clustering: thousands of similar cases-like land acquisition disputes-can be grouped so a single ruling resolves an entire set. That reduces duplication and improves consistency.
Inside the Supreme Court's SU-PACE Pilot
The Supreme Court has begun piloting SU-PACE, an AI-driven tool that reads case files, extracts issues, and flags relevant precedents. Think of it as a digital research assistant that prepares summaries and calls out the law, without issuing any judgment. It helps judges see what matters, faster.
Risks to Manage Before Scale
Justice Manmohan cautioned against blind deployment. Hallucinated case law, algorithmic bias, and privacy risks must be addressed. The human judge remains the final safeguard-integrity, independence, and intellect cannot be delegated.
- Mandatory source-citation with linkable authorities and version control.
- Human-in-the-loop review with clear thresholds for automated outputs.
- Bias and fairness testing on representative datasets, refreshed periodically.
- Privacy-by-design: data minimization, access controls, encryption, and audit logs.
- Red-teaming and error monitoring with rollback procedures.
Tech Alone Won't Fix Pendency
Dr Lalit Bhasin, senior lawyer and SILF President, warned against over-reliance on tools. He cited identical drafts from young lawyers as a signal that independent reasoning is being sidelined. That's a professional risk.
He also pointed to the numbers: pendency has crossed 5 crore cases despite better facilitation. The deeper issues-too many, outdated, overlapping, and ill-drafted laws-keep delays in place. Use technology as a tool; it's not a substitute for the human brain.
What Legal Teams Can Do This Quarter
- Map your matter taxonomy and mark "automation-eligible" categories with bright-line rules.
- Standardize data capture: structured forms, document checklists, and validation scripts.
- Deploy AI-assisted research with citation verification and judge-facing summaries.
- Cluster similar cases to streamline drafting and reduce inconsistency.
- Set governance: approval workflows, audit trails, and periodic quality reviews.
Institutional buy-in matters. Start with narrow, low-risk use cases, measure outcomes, then expand. The goal isn't to replace judges-it's to remove drudgery so judicial reasoning can do its best work.
For reference on the Court's broader context, see the Supreme Court of India. If your chambers or in-house team is building AI capability, these curated resources can help: AI automation training.
Your membership also unlocks:
