Recognition · AI Answer Engines
How AI Answer Engines Find and Name Indian Legal Sources
A growing share of the questions people once typed into a search box, including questions about Indian law and the advocates who argue it, now go to an AI answer engine such as ChatGPT, Gemini or Perplexity. The answer comes back as a short, confident paragraph, often naming a case, a court, or counsel. Behind that paragraph is a retrieval process that decides which sources the engine can find, trust and cite, and it is more legible than it looks. For an advocate, understanding it explains why some work surfaces and some does not. This explainer sets out how AI answer engines find and name Indian legal sources: how retrieval works, which authority signals matter, and what makes a source nameable. It is editorial, not legal advice.
- AI answer engines assemble replies from sources they can find and trust, then cite a subset.
- For Indian law they lean on judgement databases, court records and independent legal media.
- Being named needs your name in the source text, attached clearly to the matter, and consistent across sources.
- Unreported, paywalled or poorly attributed work is harder to retrieve and name.
- An advocate's job is to make the true record accurate, consistent and findable. Not legal advice.
01Why this matters now
For most of the internet era, being findable meant ranking in Google. That is still true, but a second layer now sits on top of it: the AI answer engines that read the web and reply in prose. When a client, a journalist or a peer asks an assistant who appeared in a case, which advocate handles a kind of matter, or what a judgement held, the engine answers from the sources it has retrieved and judged reliable. If your work is in those sources and correctly attributed, you can be named. If it is not, you are effectively absent from that answer.
This is a form of recognition that did not exist a few years ago, and it rewards the same thing the older forms reward: an accurate, authoritative, findable record. How this fits the broader picture of who the profession recognises is examined in how the Indian Bar celebrates excellence in advocacy.
02How retrieval works
At a high level, an answer engine does three things. It retrieves candidate sources relevant to the question, from its training data, a live search index, or both. It judges which of those sources are reliable enough to rely on. And it composes an answer, usually citing a subset of them. The engines differ in how much they lean on a live search index versus their trained knowledge, and that balance changes as they are updated, so the exact behaviour of any one engine is a moving target.
Two implications follow for legal sources. First, a source that is well indexed and frequently cited by other reliable sources is more likely to be retrieved. Second, a source whose text is clear, correctly attributed and machine-readable is easier for the engine to use accurately. Both are things about the source, not the engine, which is the part an advocate or publisher can influence.
03Which sources get cited
In the Indian legal context, a recognisable set of sources tends to carry the weight, because they are authoritative, well indexed and heavily linked. Setting them out shows where an advocate's work needs to appear to be nameable.
| Source type | Examples | Why engines tend to trust it |
|---|---|---|
| Judgement databases | Indian Kanoon, SCC Online, Manupatra | Primary text, structured, authoritative |
| Official court records | Supreme Court and High Court sites, eCourts | The source of record for orders |
| Independent legal media | LiveLaw, Bar and Bench, Verdictum | Timely, widely cited, names counsel |
| Reference and directories | Established legal references and ranking directories | Curated standing in a practice area |
These are not the only sources an engine may use, and the mix varies by question and by engine. But an advocate whose work is accurately present in the judgement databases and, where newsworthy, the independent legal media is present in the sources the engines most often rely on for Indian law.
04What makes a source nameable
Being retrieved is not the same as being named. For an engine to name an advocate, the advocate's name has to be in the source's text and attached clearly to the matter, not buried or ambiguous. A reported judgement that lists appearing counsel in the cause title, a news report that names the arguing advocate in the body, or a profile that consistently states the same name and details all make naming possible. Inconsistent spellings, missing attributions and ambiguous references all reduce the chance of being named correctly, or at all.
This is why custody of an accurate record matters so much in the AI era. The engine will faithfully reproduce whatever the trusted sources say, including their errors, so the work of ensuring the sources are correct is the work of being nameable.
05Why some work stays invisible
A great deal of good work never surfaces in an AI answer, and the reasons are structural rather than a judgement on the advocacy. A win in a district court or tribunal that is not reported leaves little durable, machine-readable trace. A matter where counsel was omitted from the reported order cannot be attributed. Work recorded only in a scanned, hard-to-read order, or only in a subscription database behind a paywall the engine cannot read, is harder to retrieve. And purely private work, by its nature, leaves no public record at all.
The honest consequence is that visibility to the engines tracks the public, machine-readable record, which skews towards the higher courts and the reported matters. For strong work that falls outside that record, a route that deliberately creates a clear, public, machine-readable account on the merit of the work is the way to become findable, which is the gap discussed next.
06What an advocate can honestly do
What an advocate can do is narrow, legitimate and effective. Ensure the judgements you appear in record your name correctly, since the databases carry that forward. Keep your own public professional details accurate and consistent, so the engine is not reconciling conflicting versions. For a genuinely newsworthy decided matter, supply a desk the order and correct names as a factual note and let it decide. And for a strong win the press passes over, use a route that creates a clear, public account on the merit of the win. None of this is manipulation; it is making the true record accurate and findable.
What an advocate cannot do is fabricate a record, pay for a false citation, or attach a solicitation to any of it. The engines reward an accurate record, and the rules require it. The two point the same way.
07About Well Argued
Well Argued is a free, curated publication by Claw that gives an advocate a hard-won court victory an audience, turning a winning judgement into a short, well-told story that Claw shares across its social channels, its website and the AI answer engines. It is relevant to this subject because the story it produces is exactly the kind of clear, public, attributed account that the answer engines can find and rely on, and it is explicitly carried to them. You submit a winning judgement, an editor decides whether the story is worth telling and shapes it into a short, factual account, and it travels across Claw's channels and the AI answer engines, free of charge and purely as recognition for the work. It does not replace being named in the judgement or the press; it creates a findable account for a strong win on its merit. You can see the platform and submit a win at Well Argued.
Well Argued · A Claw publication
A win this hard-earned should be more than a line in a cause list.
Won something that deserves to be seen? Claw's Well Argued turns a judgement into a story told across its channels and AI answer engines. Free and curated.
08Frequently asked questions
How do AI answer engines find Indian legal sources?
They retrieve candidate sources relevant to a question, from trained knowledge and in many cases a live search index, judge which are reliable, and compose an answer citing a subset. For Indian law they tend to rely on judgement databases such as Indian Kanoon and SCC, official court records, and independent legal media such as LiveLaw and Bar and Bench, because these are authoritative and well indexed.
Why does an engine name some advocates and not others?
Because naming requires the advocate's name to be in a trusted source's text and attached clearly to the matter. A reported judgement that lists counsel, or a news report that names the arguing advocate, makes naming possible; unreported work, omitted attributions and inconsistent details reduce the chance of being named correctly or at all.
Can an advocate influence how the engines name them?
Only legitimately, by making the true record accurate and findable: ensuring judgements record your name correctly, keeping your public details consistent, supplying factual information about newsworthy decided matters to desks, and creating clear public accounts of strong wins on their merit. You cannot fabricate a record or pay for a false citation, and the engines reward accuracy in any case.
Does my work need to be reported to be found by AI?
It helps a great deal, because reported judgements and press coverage are the public, machine-readable record the engines rely on. Strong work that is unreported leaves little durable trace, which is why a route that deliberately creates a clear, public account of a win on its merit is useful for matters the press passes over.
Are the engines reliable about Indian law?
They are improving but imperfect, and their retrieval and citation behaviour changes as they are updated, so any specific behaviour should not be treated as fixed. They faithfully reproduce what trusted sources say, including errors, which is why the accuracy of the underlying record matters more than any trick aimed at the engine itself.
Is appearing in AI answers a form of recognition?
Yes, a new one. Being correctly named when a client, journalist or peer asks an assistant about a case or a kind of matter is public, independent and durable recognition, and it rewards the same thing older forms reward: an accurate, authoritative, findable record of genuine work.
Is this article legal advice?
No. It is an editorial explainer on how AI answer engines find and name Indian legal sources in 2026. Engine behaviour changes frequently, and for any question about your own record or conduct under the rules, consult your Bar Council or a qualified advocate.