AI fix for India’s accident tribunals

On November 23, 2022, Arunkumar, a 30-year-old bank manager from Hubballi, died in a road accident while returning home. He left behind his wife and their one-year-old son. Four months later, his dependants filed a petition before the district Motor Accident Tribunal. It took the Tribunal over 18 months to decide compensation – and it undervalued the award by nearly Rs 50 lakh.

Arunkumar’s case is one of almost nine lakh accident claims pending before tribunals and high courts nationwide. Nearly three lakh of these are before high courts alone. The system is clearly failing victims.

Motor accident compensation involves little more than straightforward calculations, besides witness depositions and lawyers’ submissions. Yet cases take an average of four years to conclude. Justice is delayed not only
for accident victims but for millions of other litigants, as motor accident cases consume almost half the working hours of senior civil courts. In 2023, Justice Arvind Kumar urged the use of artificial intelligence to determine compensation.  

The Supreme Court introduced the Electronic Detailed Accident Report (E-DAR) scheme through orders in WP 295 of 2012, with the aim of disposing of cases within four months of the police reporting accidents to tribunals and insurers. Despite repeated directions and close monitoring by the Supreme Court Committee on Road Safety, the scheme has achieved only limited success – even in Delhi, the best-performing state. Disposals there barely outpace new filings. Elsewhere, the impact is negligible.

Why has the scheme faltered? Largely because stakeholders navigate the process poorly. Police treat E-DAR as a mere paperwork exercise -- sending documents to tribunals, insurers, and legal service authorities – while failing to inform victims of their rights. This is why most E-DARs (except in Delhi) do not translate into tribunal proceedings, contrary to the SC’s intent.

Insurers, legal service authorities, and tribunals act as passive recipients of E-DARs, with minimal follow-up until a victim or lawyer files a petition. The four largest insurers, who control 30-40 per cent of the market, lack internal guidelines for handling E-DAR notices. The scheme imposes little responsibility on insurers or tribunals despite their central role in timely disposal. Tribunals are not even part of the integrated road accidents database, which provides real-time information on accidents and victims.

Although the law treats motor accident cases as inquiries –not trials – lawyers continue to handle them like regular civil suits. Supreme Court directions require them to file written statements in specific formats meant to make proceedings cooperative. These directions are routinely ignored, resulting in disorderly hearings and legally flawed judgements. Nearly a third of all pending cases are challenged before high courts, mainly arising from tribunal-level errors.

The problems compound at the high court level due to calculation or judgement errors by tribunals and frivolous appeals. Many such appeals are later settled in the Lok Adalats for as little as Rs 15,000-20,000—often less than the legal fees involved. 

Digital solutions can resolve many of these problems. First, tribunals need automated tools to follow up with police for missing documents and to issue digital summons to victims and insurers – kickstarting proceedings without delay.

Second, lawyers’ initial submissions should be filed through digitised versions of the SC-prescribed formats. This would combine multiple early stages, such as framing issues, application hearings, and written statements, into a single stage. Such simple digitisation could cut pendency by 17-20 per cent, while helping tribunals frame issues correctly and determine questions such as burden of proof with clarity.

Most cases require depositions only from eyewitnesses and investigating officers. But tribunals routinely summon insurers and doctors merely to reiterate admitted facts. This needless practice makes witness examination account for nearly 40 per cent of pending cases in Karnataka and 24 per cent nationally. Kerala avoids this and keeps the stage to just 6 per cent of its backlog. Mandating digital submissions—including admission or disputing of relevant facts—would end the conversion of these enquiries into adversarial trials. This would align with the statute’s spirit reaffirmed by the Supreme Court in Srikrishna Kanta vs Oriental (March, 2025).

Once initial stages are digitised, an AI tool trained on accurate legal datasets can generate draft judgements based on admitted facts. Lawyers may accept or contest these drafts, and tribunals can issue final orders after review. Combined with AI tools, these reforms can deliver faster, accurate judgements --removing the need for lengthy litigations to perform basic calculations. Civil courts would then be free to focus on complex disputes where human reasoning is essential.

(The writer, a graduate of the National Law School of India University, Bengaluru, is a legal officer with a general insurance major) 

Disclaimer: The views expressed above are the author's own. They do not necessarily reflect the views of DH.