Can AI Fix India’s Justice Delays?
With crores of cases stuck in India’s courts, AI offers hope. From Lawnet to SUPACE, can technology make justice faster and fairer?
Verdicto News Desk 
India’s justice system is burdened by over 5 crore pending cases, making the delays chronic and costly. Singapore’s AI-powered platform Lawnet, which allows legal professionals to ask questions in natural language and receive instant, sourced responses from case law and legislation, offers a possible model. But can India build something similar that works at its scale and respects its constitutional values? Below, along with policy perspectives, are insights and concerns from Indian academics, judges, and innovators.
What Judges and Legal Experts Have Already Said
Before turning to academic voices, a few judicial and professional quotes:
- Chief Justice of India D.Y. Chandrachud has called AI in legal research a “game-changer”, emphasising its potential to boost efficiency and accuracy. But he also warned of the need for ethical guardrails—on transparency, fairness, and data bias.
- Supreme Court Justice Sanjay Kishan Kaul has acknowledged AI’s usefulness for legal research and drafting, but stressed that human qualities—judgment, empathy, wisdom—remain essential. He said:
“A judge operates from the mind and the heart. I don’t believe AI can take care of the latter aspect.”
- Supreme Court Judge Manmohan urged that India’s legal system must evolve to govern emerging tech responsibly. He pointed out challenges around privacy, cybersecurity, IP rights, especially in crossborder contexts.
These voices show optimism moderated by caution, AI is seen as a tool, not a substitute for human judgment.
Here are insights from law professors, researchers, and legal tech innovators, based on recent studies and interviews:
- Evaluating LLMs’ Performance in India
In a recent empirical study, researcher Rahul Hemrajani evaluated how Large Language Models (LLMs) perform on key Indian legal tasks—issue spotting, drafting, research, reasoning—comparing them with law students and junior lawyers. The findings were mixed: LLMs performed well in drafting and spotting issues, often matching or even surpassing students in certain metrics. But they were weaker in specialised legal research, and prone to “hallucinations” (producing outputs that are factually incorrect or citing non-existent precedents). The study underscores that LLMs can assist, but cannot replace, human expertise. - Academic Concerns in Legal Education
Professors are increasingly worried about misuse and over-reliance on AI in law schools. A recent article in The Daily Guardian noted that challenges include: plagiarism risks, erosion of critical thinking if students just accept AI-generated output; lack of institutional guidelines; limited awareness among faculty; infrastructure gaps (internet, computing resources).
One piece argues legal pedagogy should evolve: integrating technology not just as tools, but in curricula that teach students how to evaluate AI output, how to question assumptions, how to recognize bias or error.
- AI Accessibility & Tools
- The Bar Council of Delhi (BCD) has set up an e-research library in the Rouse Avenue courts with access to legal databases (Manupatra, SCC Online) and an AI tool called Lucio, to help lawyers speed up research. Frequency: lawyers get access via desktops; more remote access is being planned.
- Initiatives like Adalat AI (speech-recognition/transcription of court proceedings) are helping reduce load in procedural tasks. Projects like these show how AI can be introduced gradually, starting with supportive roles.
- Regulatory, Ethical & Governance Issues
Researchers note that India lacks a comprehensive legal framework or binding guidelines about AI in law practice: how to validate sources, liability if AI gives wrong research, disclosure obligations when lawyers use AI, data privacy concerns.
Academics also point out that without checks, AI systems may reproduce historical biases in judgments/statutes, disadvantaging marginalized groups. Ethical AI must include fairness, explainability, and human oversight.
Policy & Design Recommendations
Drawing on what judges, scholars, and innovators are saying, here are refined policy proposals for an AI research platform India should consider:
|
Area |
Recommendations |
|
Ethical Framework & Transparency |
Legally binding guidelines about source validation, dealing with hallucinations, data provenance; mechanisms for audit; disclosure of confidence or uncertainty in AI’s responses. |
|
Human-in-the-Loop |
Every AI output must be vetted by lawyers / judges; AI to assist not decide; ensure human judgment, equity, empathy are preserved. |
|
Regulation & Accountability |
Define who is liable if AI output misleads—tool developers, lawyers using them, or platform providers; bar councils / judicial bodies to set standards. |
|
Legal Education Reform |
Law schools need to teach critical use of AI: how to cross-check sources, question AI reasoning, understand limitations. Curricula should include legal tech and ethics. |
|
Infrastructure & Access |
Uniform digital records, reliable internet & device access; access for rural / smaller courts; open databases of judgments with metadata. |
|
Pilot Programs |
Start with use cases like legal research aid, translation, transcription (speech-to-text), and natural-language query tools; monitor outcomes (speed, accuracy, user satisfaction). |
|
Language & Inclusivity |
Tools must support Indian legal terminology, statute language, judgments from various states; translation into regional languages; accessibility for persons with disabilities. |
|
Data Privacy & Bias Mitigation |
Strong rules around personal data used in judgments; techniques to detect/mitigate bias; ensure diverse training data. |
AI can’t erase India’s judicial backlog overnight but it can be one of the strongest levers to reduce delay and increase access to justice
India has the talent (judges, legal scholars, tech innovators), the need (huge pendency, procedural delays), and some early pilots (transcription, e-libraries). What’s required is a concerted push to build a platform (or multiple interlinked tools) that:
- are rooted locally (use Indian cases, statutes, regional/vernacular languages),
- are transparent and accountable,
- preserve human judgment and values,
- and come with the regulatory and ethical guardrails to prevent misuse.
If done well, such a platform modeled along the lines of Lawnet, but adapted to India’s scale, diversity, and constitutional commitments, could help reduce pendency in meaningful ways, speed up the delivery of judgments, and improve legal literacy among both lawyers and litigants.
📌 Follow us on YouTube, Instagram, and Twitter for more updates.
