Methodology

Where the material on MY Case Law comes from, how it is put together, what that process gets wrong, and how to tell us when it has.

Primary sources only

Everything on this site starts from written judgments released by the Malaysian Judiciary (Kehakiman Malaysia) through eJudgment, the Judiciary's official judgment portal, and from Malaysian legislation as published in Malaysian Legislation (lom.agc.gov.my). We add no facts from anywhere else — not from news reports, not from directories, not from practitioners' own accounts of their cases. Every case page links back to its source so the judgment can be read directly rather than taken on our word.

eJudgment publishes a subset of Malaysian judgments, not all of them, and the portal's coverage starts well after the courts did. A case that is absent from this site is not a case that was never decided — it is, far more often, one that was never written up for publication or falls outside the period the portal covers. Nothing here should be read as a complete picture of what any court, judge or advocate has done.

How pages are produced

This site is built with AI assistance, and we would rather say so plainly than imply otherwise. The judgments are the courts' own words and are not rewritten. Everything around them — the summaries, the explainers, the narrative text on a judge, counsel, firm or court page — is drafted by AI systems working from the judgment text, and then loaded to the site. Structured facts such as court, decision date, parties and counsel are extracted from the judgment rather than written freehand.

Three limits on that, stated up front rather than at the bottom:

• No advocate and solicitor reviews these pages. There is no qualified Malaysian practitioner checking the prose before it is published, and we do not pretend there is by attaching a lawyer's name to it. Pages are published under organisational authorship, described below.
• AI-drafted text can be wrong. It can misread a holding, flatten a distinction that mattered, or state something with more confidence than the judgment supports. Where our summary and the judgment disagree, the judgment is right and we are wrong — which is why every page links to the source.
• Some judgments are held only as records. Not every case in our index has usable full text behind it. Where that is so, the page carries the case's citation, court, date and parties and points to eJudgment for the judgment itself, rather than inventing an analysis of a document we cannot read.

How entities are derived

Pages about judges, counsel, law firms, courts and organisations are not curated by hand. They are derived automatically: names are extracted from the judgment text — from the coram, the appearances and the parties — normalised, and grouped into entities, which is then what a judge's or an advocate's page is built from.

That process is imperfect, and in predictable ways. Malaysian judgments spell the same person several ways across different courts and years — with and without honorifics, with initials expanded or not, in different transliterations — so one person can end up split across more than one page, or two people with similar names can be merged into one. Firm names appear in long form, short form and abbreviation. A misattributed appearance follows the misattribution wherever the judgment recorded it. We would rather describe these failure modes than let a reader assume a page is an authoritative record of someone's practice. It is a view of what the published judgments happen to say, with the errors that come with reading thousands of documents automatically.

One category of page is deliberately not built at all: individuals named as parties. Those entities are never indexed and never appear in any sitemap, at every case count and every stage of the site's rollout. The reasoning is set out in full on our privacy policy.

How practice areas are derived

Practice-area labels are derived from each judgment's catchwords — the subject headings the court or the publisher attaches to a decision — together with the statutes and terms the judgment itself uses. They are a classification we compute, not a label assigned by a lawyer, and they are not part of the court's decision.

This is the weakest derivation on the site, and the honest description is that a meaningful share of the corpus is not classified at all. Catchword formatting varies widely across Malaysian courts and registries, judgments are written in both Malay and English, and the same area of law appears under several different headings. Where our classifier cannot recognise a heading with reasonable confidence, the case is left unclassified rather than pushed into an approximate bucket — so an area page shows cases we could identify, never all the cases that exist on that subject. Improving this coverage is active work; until it lands, a practice-area page is a starting point for reading, not a survey.

Organisational authorship

Pages are compiled and published under organisational authorship, as the MY Case Law editorial team. We do not attribute pages to a named lawyer, because none wrote them — inventing a byline would be the easiest way to make this site look more authoritative than it is, and it is the first thing we decided not to do. Malaysia's legal profession is fused, and an advocate and solicitor is the person qualified to give advice on a matter; this site is not that, and does not stand in for it.

What we don't publish

• We don't rank or score judges, advocates and solicitors, or firms, and we publish no win rates, success rates or head-to-head comparisons. A case record is not a performance measure, and treating it as one would misrepresent both the profession and the corpus.
• We don't publish predictions about how a court would decide a matter, or material framed as guidance on what to do about one. Where a page would have to reason past what the judgments actually say, it is not published.
• We don't republish the full text of judgments as our own content. The court's version on eJudgment is the authoritative one, and that is where we send readers for it.
• We don't build profile pages about individuals named as parties — see above.

Corrections & data removal

Given everything above, errors on this site are not a hypothetical — automatic derivation and AI-drafted prose both produce them. This is the visible channel for reporting one, and for asking that personal data drawn from a judgment be corrected or removed. It is also the route for a data access or data correction request under Malaysia's Personal Data Protection Act 2010 (Act 709), which our privacy policy describes in more detail.

• What can be reported. A page that misstates what a judgment says; a case attributed to the wrong judge, counsel or firm; one entity that is really two people, or two that are really one; a mislabelled practice area; a wrong date, court or citation; or personal data appearing on a page that should not be there.
• How to report it. Email [email protected] with the URL of the page and what specifically is wrong. The case citation, and a pointer to the passage in the judgment, let us verify it against the source quickly. There is no charge and no form to fill in.
• What happens next. We check the report against the judgment itself. Where the page is wrong we correct it, or remove it where correction is not the right remedy, and the page's "Updated" date moves to reflect that a real change was made — we don't bump dates cosmetically. Where the page turns out to match the judgment, we say so and explain what we found, rather than leaving the report unanswered. We aim to acknowledge reports and act on them within a reasonable period.

One limit worth being clear about: written judgments are public records released by the Malaysian Judiciary, and the authoritative version of any judgment stays with the court's own portal regardless of what we change here. Where the objection is really to what a judgment says or records, the correction that matters is the court's, not ours — and we will say so plainly rather than let a request go quiet.

Freshness

Judgments are appealed, distinguished and overruled, and statutes are amended. A page on this site reflects the judgments we held when it was compiled, and it does not automatically know that a decision it describes has since been reversed on appeal. As new judgments are loaded we re-derive the pages they affect and show an honest "Updated" date when the content actually changed. For whether a case is still good law today, the judgment and the later authorities are the record; this site indexes what was decided, and does not track its subsequent treatment.