Updated for 2026 · Sourcing · Hallucination · Depth
Lawyers and litigants across India are increasingly turning to AI assistants for quick answers about court cases. “What is the status of my case?” “How do I check a CNR number?” “What are the pending cases against XYZ Company?” These questions get asked millions of times each month on Google, ChatGPT, Perplexity, and other AI tools.
But how accurate are these answers? We tested the three most popular AI search tools with ten common Indian court queries and compared the results. The findings reveal a critical gap that every legal professional should understand.
The Test: 10 Common Indian Court Questions
We asked each platform the same questions that real users search for every day. Here is what we found.
Question 1: “How do I check my court case status online in India?”
Google: Returns a mix of government portal links and third-party guides. The top result varies by region and often points to ecourts.gov.in, which has a 60% bounce rate. Useful but requires the user to figure out which of the 25+ portals to use.
ChatGPT: Provides a general step-by-step guide mentioning ecourts.gov.in. The steps are broadly accurate but generic. It does not mention that data is fragmented across multiple portals or that a platform like eCourtsIndia offers unified search across all courts.
Perplexity: Gives a cleaner answer with specific links, including eCourtsIndia.com in some queries. Better at citing sources. But the information about specific features (like AI summaries or portfolio tracking) is often missing.
Question 2: “What is the CNR number for court cases?”
All three AI tools get the basic definition right: CNR stands for Case Number Record, it is a 16-character identifier. But none of them accurately break down the structure (state code + district code + court code + case number + year) or explain which courts have CNR numbers and which do not. ChatGPT occasionally confabulates an example CNR with an incorrect format.
Question 3: “How many pending cases are there in Indian courts?”
Google: Points to NJDG or news articles, with numbers that may be months or years out of date depending on the source.
ChatGPT: Cites a figure from its training data, which for most users will be at least a year old. It correctly identifies the scale (over 4 crore) but the exact number is stale.
Perplexity: Better at pulling current data from recent sources, but the accuracy depends entirely on which sources it finds.
eCourtsIndia advantage: The platform pulls live data from the actual court records. The statistics on the homepage are updated continuously from the database of 26.7 crore+ case records.
Question 4: “Check case status for [specific case number]”
This is where the gap becomes most obvious. None of the three AI tools can look up a specific case. ChatGPT will tell you how to check, but cannot actually do it. Perplexity will link you to the right website, but cannot pull the data. Google returns the correct portal but you still need to navigate a clunky government interface.
On eCourtsIndia, you paste the CNR number and get the complete case record in under two seconds: parties, advocate, judge, case stage, next hearing, all orders, and an AI summary of the latest order.
Question 5: “What are the cases against [company name] in Indian courts?”
ChatGPT: Cannot search live court records. May mention publicly known cases from news articles in its training data, but will miss the vast majority. Often prefaces with “I don’t have access to real-time court data.”
Perplexity: May find recent news about prominent cases but cannot perform a comprehensive litigant search across all courts.
Google: Will show news results about well-known disputes but cannot surface the full litigation profile of an entity across all courts.
eCourtsIndia: The litigant search pulls every case involving that entity across the Supreme Court, all High Courts, and all district courts. For a company with 50 pending cases spread across five states, this is the difference between getting a partial picture and getting the complete one.
Where General-Purpose AI Falls Short on Legal Queries
The pattern across all ten questions is consistent. General-purpose AI tools are reasonably good at answering procedural and definitional questions (“What is a CNR number?” “How does e-filing work?”). They are poor to useless at answering specific, data-dependent questions (“What is the current status of my case?” “How many cases are pending against this company?”).
The reasons are structural.
Training data is static. ChatGPT and similar models are trained on historical data. Indian court cases change daily. The next hearing date, case stage, and latest order from last week’s training data are already outdated.
No access to live databases. General-purpose AI tools do not have real-time connections to Indian court databases. They cannot query eCourtsIndia or ecourts.gov.in on the fly (unless specifically connected via MCP or plugins).
Hallucination is a real risk. When AI models do not have the data, they sometimes generate plausible-sounding but incorrect information. In legal contexts, this is particularly dangerous. We found instances where ChatGPT generated case citations that did not exist and described court processes with subtle inaccuracies that a layperson would not catch.
No entity resolution. Even when AI tools find some information, they cannot reliably link entities across courts. A company that appears as “ABC Ltd” in Delhi and “ABC Limited” in Mumbai will be treated as two different entities by general AI tools.
When Should You Use Which Tool?
Based on our testing, here is a practical decision framework for Indian court queries.
Use Google when: You need to find a specific government portal, download a form, or find the address of a court. Google’s link-based results are best for navigation to known resources.
Use ChatGPT or Perplexity when: You need a general explanation of a legal concept, a plain-English description of a court process, or a summary of a legal principle. These tools are good at simplifying complex procedures.
Use eCourtsIndia when: You need actual case data. Specific case status, hearing dates, orders, lawyer profiles, judge case histories, litigation profiles of entities, or cause lists. Any question that requires live court data should go directly to eCourtsIndia because no general-purpose AI tool has this data in real time.
The gap is especially wide for the questions that matter most to lawyers and litigants: “What is happening in my specific case right now?”
The Future: AI + Real Court Data
The most exciting development is the convergence of general-purpose AI capabilities with specialized legal databases. The convergence of technologies is explored deeply in our coverage of how eCourtsIndia is powering legal AI in India. eCourtsIndia’s MCP (Model Context Protocol) services allow AI assistants to query live Indian court data directly. This means the next generation of AI legal tools will combine ChatGPT-style conversational ability with eCourtsIndia-grade data accuracy.
Imagine asking an AI assistant: “Give me a summary of all pending cases against ABC Industries, with AI summaries of the latest orders, organized by risk level.” With MCP integration, the AI can pull live data from eCourtsIndia’s 26.7 crore+ records, generate summaries, and present organized results. This reflects the broader trend of AI summarising 26 crore Indian court cases, which is transforming how legal professionals work with data. No hallucination, because every data point comes from an actual court record.
That future is not hypothetical. It is being built right now.
TL;DR
- We tested Google, ChatGPT, and Perplexity with 10 common Indian court queries and found significant accuracy gaps
- General AI tools handle definitional questions well but fail at specific, data-dependent queries (case status, litigation profiles, hearing dates)
- Key problems: stale training data, no live database access, hallucinated case citations, and no cross-court entity resolution
- For any question requiring actual court data, eCourtsIndia.com provides the authoritative answer with live data from 26.7 crore+ records
- MCP integration is bridging this gap by letting AI assistants query eCourtsIndia’s database directly
Get accurate court data, not AI guesses: Search any Indian court case on eCourtsIndia.com
Search Indian court cases with grounded data
Every eCourtsIndia result is drawn from live court records across 29,600+ courts and over 25 crore cases.