AI-based resolution of cheque bounce cases is feasible: Ex CJI DY Chandrachud
Ex Chief Justice of India DY Chandrachud argued that India can deploy artificial intelligence to decide limited, high-volume disputes-most notably cheque dishonour matters. The goal: clear pendency without trading away fairness. The message was delivered at the IBA Litigation and ADR Symposium during a keynote on technology, constitutionalism, and the future of dispute resolution.
Why cheque dishonour cases fit an AI model
Cheque bounce litigation is large, standardised, and transactional. Elements are patterned, evidence is document-heavy, and outcomes rarely implicate core rights in the same way as housing or liberty questions do.
Justice Chandrachud pointed to Delhi's virtual courts as proof that automation can safely absorb routine work and free judges for matters that demand full adjudication. An AI-led track for Section 138 cases under the Negotiable Instruments Act, 1881 could follow the same logic, with standardised inputs, structured outputs, and clear options to contest.
For reference, India has already run traffic matters through digital rails with Virtual Courts, reallocating judicial time to higher-stakes disputes. The same architecture can be adapted for cheque disputes with bank data integrations, template orders, and secure service.
Where human oversight stays essential
Housing and rent control disputes must remain under a judge's eye due to the risk of eviction and displacement. These cases carry dignity interests that an automated system should not decide.
He also flagged motor accident compensation as a candidate for optional AI. Insurers could be bound by instant, rules-based awards while victims retain a choice: accept the offer or seek full judicial adjudication.
Efficiency as a constitutional value-without eclipsing fairness
Justice Chandrachud underscored that efficiency cannot be treated as a luxury anymore. If the system remains slow, those who need relief most do not get it in time.
At the same time, he cautioned against letting efficiency erase access, contestation, or dignity. His provocation-accepting "some degree of injustice" to avoid systemic breakdown-was a warning about trade-offs, not a license to cut corners. Any AI rollout must preserve the right to be heard where rights are at stake.
What this means for legal teams
- Litigators: Prepare evidence packs for Section 138 in machine-readable formats; expect tighter timelines and portal-driven filings; build playbooks for contesting automated findings on narrow grounds.
- In-house counsel: Tighten cheque issuance controls, consent workflows, and audit trails; anticipate faster demand cycles and earlier settlement decisions; set escalation rules for disputed AI outcomes.
- Claims and defense teams: For motor accidents, model early-value bands, reserve strategies, and acceptance thresholds if optional AI awards become available.
- Court administrators: Prioritise explainability, audit logs, and appeal layers; publish error taxonomies and correction windows to protect legitimacy.
Implementation checklist for policymakers
- Define scope narrowly: start with uncontested or low-discretion Section 138 matters; phase in complexity.
- Consent and choice: default to AI for routine steps, but keep a clear path to a human judge on request.
- Explainability: every decision must expose inputs, rules applied, and reasons that can be challenged.
- Data plumbing: API links with banks for cheque images, return memos, and account confirmations; standard e-service and acknowledgment.
- Safeguards: detect identity fraud, coercion in settlements, and systemic bias across geographies and languages.
- Remedies: fast-track review, simple set-aside procedures, and strict timelines for escalation.
- Pilot first: time-bound pilots with public metrics (time-to-disposal, error rates, user satisfaction), independent audits, and sunset clauses.
Risks to manage
- Black-box decisions erode trust. Use transparent rules, open documentation, and reasoned outputs.
- One-size-fits-all penalties. Keep proportionality checks and a mechanism to account for context.
- Digital exclusion. Offer assisted filing centers, multilingual interfaces, and offline-to-online bridges.
- Due process drift. Preserve notice, response rights, and a meaningful opportunity to contest.
What to watch next
- Pilots in magistrate courts for Section 138 with opt-out rights and limited, reviewable AI orders.
- Optional AI awards in motor accident claims that bind insurers while protecting claimant choice.
- Standard templates for reasons, error correction, and data-sharing protocols with banks and law enforcement.
- Rules or practice directions clarifying status, enforceability, and appealability of AI-generated orders.
The takeaway for the bar and bench is simple: use AI where the law is clear, facts are structured, and stakes are bounded. Keep humans where dignity, housing, liberty, or complex causation are in play. That balance can move dockets without hollowing out justice.
If you're building internal capability to evaluate AI tools for dispute resolution, here's a curated hub of courses by job role: Complete AI Training.
Your membership also unlocks:
