As artificial intelligence (AI) continues to evolve, its potential integration into various sectors has sparked debates worldwide. One area drawing increasing attention is its use in the judiciary.

Saddled with the backlog of millions of pending cases, the Indian judicial system might benefit from using artificial intelligence, experts said. However, they caution that such a tech-led system would miss some of the unique nuances of cases that require human judgement.

Tony Blair Institute for Global Change (TBIGC) country head Vivek Agarwal, speaking with Mint, said AI features in judiciary could significantly reduce case backlogs and improve efficiency, citing Japan’s ongoing use of AI to draft judgments in procedural matters as a success story.

Former UK prime minister Tony Blair’s eponymous non-profit recently said it is advising a few state governments in India to consider using AI for administrative purposes within their offices. 

Agarwal pointed out challenges specific to India, such as algorithmic bias based on caste or gender, language barriers, lack of technology, inconsistent data formats, and judges’ reluctance to adopt AI.

High courts in Delhi, Karnataka, Telangana, Punjab, and Haryana have been experimenting with AI tools for administrative reasons.

The Supreme Court of India alone has 66,054 pending cases. Add to that over 6 million cases in the high courts and a multitude of others in the lower courts. Several legal experts suggested using AI to improve administrative efficiency and even handle complex tasks like reviewing evidence.

In sectors like insurance, banking, and e-commerce, AI tools could serve as interactive guides, helping litigants understand laws through dynamic frequently asked questions (FAQs), they said.

The Indian judiciary is presently integrating AI for legal research, translation and predictive justice. In 2021, the apex court launched Supreme Court Portal for Assistance in Courts Efficiency (SUPACE) to assist judges with relevant case laws and precedents and Supreme Court Vidhik Anuvaad Software (SUVAS), an AI tool to translate judgments and orders into nine regional languages.

A bunch Indian datacentric AI companies are actively working in the judicial space to offer AI solutions. Nyaay AI, a product of Mumbai-based Indika AI provides solutions like automated filing by extracting data from case documents, detection of defects in the filings and smart case triaging and bunching to determine urgency. The system also helps in judgment research and machine translation.

Hardik Dave, founder of Indika AI, said the company was using proprietary or open-source AI platforms locally to maintain full control over judicial data to maintain privacy. It is also deploying AI on government servers to ensure data localization and compliance and data anonymization to redact sensitive information before processing to protect privacy. He said that training would be crucial to know how AI tools integrate into existing systems. “Judicial staff and lawyers need to learn AI basics, data privacy, and ethical use to ensure responsible AI application,” he said.

Future use of AI

Lawyers see potential in using AI for certain specific areas. Mint reported in March how large Indian law firms are moving rapidly to adopt artificial intelligence to carry out a string of ‘simple’ tasks including research, drafting and client presentations, helping them achieve efficiency and leaving the lawyers to focus on litigation.

Suvarna Mandal from Saikrishna and Associates said she sees promise in AI for automating case scheduling through sophisticated algorithms, drafting administrative and procedural orders for final review by judges, flagging cases for alternative dispute resolution, and managing case files. 

“AI can have potential use through intelligent chatbots or legal query systems accessible to the public that could provide guidance and basic legal information to help litigants,” she added.

Over-reliance on AI outputs can influence judges’ rationality, leading to biased decisions.

Others are even more optimistic and believe AI can take on more complex tasks that currently require human intervention. Justice Gautam Patel, a former judge of Bombay high court, said AI-driven models can detect patterns in legal decisions, assist in policymaking and aid in case management. For example, AI could map trends in case distribution based on geography or age, providing policymakers with data to allocate resources better, such as increasing the number of judges in overburdened courts, he said.

The former judge explained that AI’s ability to analyse judicial trends could help lawyers and litigants make informed decisions by identifying how different judges typically rule in specific areas of law. By offering data-driven insights, AI could ensure that legal strategies are more tailored and efficient, ultimately aiding in faster case resolution.

Also Read: Open Network courts can enable vast judicial process reforms

Limits and risks of AI

While AI’s potential in the judiciary is clear, it also faces criticism. Experts argue that AI, lacking the expertise of legal practitioners, fails to account for ethical standards and legal requirements.

Former chief justice of Bombay and Rajasthan high courts Pradeep Nandrajog and former vice president of Supreme Court’s e-court committee justice R.C. Chavan emphasized AI’s limitations in adjudication, noting that each case is factually unique and cannot be reduced to patterns or past precedents alone. They argued that AI cannot account for human factors like corruption, omissions, or subjective nuances common in litigation, especially in India.

Chavan pointed out that AI relies on precedent, which could slow the evolution of legal principles. He questioned whether legal outcomes should be shaped by rigid algorithms anchored in the past. Nandrajog also raised concerns about AI’s inability to handle culturally nuanced cases, such as phonetic trademark disputes, where regional pronunciation differences could impact outcomes—something AI might not interpret correctly.

Some are worried that overreliance on AI could lead to a host of challenges. “Over-reliance on AI outputs can influence judges’ rationality, leading to biased decisions. Judges, meant to uphold equality and neutrality, may develop a preference for AI’s suggestions, resulting in judgments that contradict the fundamental rights of the citizens,” Akshat Pande, managing partner at Alpha Partners said.

Striking a balance

As AI technology advances, the pressure to adopt it within the judiciary will likely grow, especially in overburdened systems like India. Justices Patel and Chavan clarified that while AI is essential for the judiciary, it should not dictate judgments. They said that AI could help streamline court operations, but the substantive adjudication of cases should remain firmly in human hands.

Companies promoting AI also agreed that courts should not try to replace human judgment altogether. “AI will assist judicial processes, but fully automated verdicts remain unlikely, especially for complex cases. However, AI could handle minor, non-contentious cases where quick resolutions are needed,” Dave said.