What the Legal Information Institutes Hold

The phrase has a published definition, which is unusual for a label that spread as widely as this one. Institutes meeting in Montreal in 2002 declared what they were undertaking to do, and the declaration has been amended at four later meetings, the last of them in 2012. Anyone deciding whether to rely on a service calling itself a legal information institute can start by reading the four things the name is supposed to mean.

They are: publish public legal information originating from more than one public body, over the internet; provide free and anonymous public access to it; do not impede anyone else from obtaining the same material from its sources and publishing it; and support the objectives the declaration sets out. Nothing in that list is about comprehensiveness, currency, or whether the text served is the one a court would use. One of the four is about scope, one about access, one about not standing in anybody else's way, and the last is a statement of good faith.

Read carefully, the definition tells a researcher less than the label seems to promise, and something the label does not hint at.

A state gazette can pass all four tests

Free and anonymous access is the criterion that sounds distinctive, and it is not. Austrian law puts that same duty on the state, requiring the contents of its federal gazette to be reachable at all hours, for nothing, with nobody asked to identify themselves. It is more than the declaration asks, and it is asked of a government rather than of a volunteer project.

Nor does the scope test separate them, on a careful reading. A national gazette carries instruments made by a legislature, by a head of state and by several ministries, which is already information originating from more than one public body. Take the four tests strictly and they do not draw a line between an institute and a state publisher at all. That is easy to read as loose drafting. The case below suggests it is not. The alphabetical list of the jurisdictions covered here is a fair sample of how differently the two can behave while passing the same four tests.

The label does not tell you who is behind it

The clearest illustration is Kenya Law, which sits in the African network's own list of institutes. Its organisational profile describes it as the brand name of the National Council for Law Reporting, a semi-autonomous state corporation, whose council is chaired by the Chief Justice, and whose statutory mandate is to report the development of the country's case law and to revise, consolidate and publish the laws. The same page describes it as the official publisher of those two things, and states its membership of the movement several sections further down, under its policy statements.

That is a state publisher and a legal information institute at once, and the two facts sit a few paragraphs apart on one page. A researcher who has absorbed the idea that an institute is a third party republishing what a government produced, and that a government publisher is the place to go for the authentic text, has no box for it. The lesson is not that the categories are useless. It is that the name on the site is not evidence of which one applies, and the about page usually is.

Holding and pointing are different things

The other ambiguity in the word holds is between a collection a service publishes and a list of addresses it maintains. The institute at Cornell Law School states its own position on this plainly: its collection of world legal materials gathers, country by country and continent by continent, the internet-accessible sources of foreign constitutions, statutes and judicial opinions, and it holds resources and document collections of international law alongside them. Gathering the sources is not the same as holding the texts, and the distinction is the difference between a service whose contents can go stale and one whose links can go dead.

Both failure modes are real and they need different checks. A stale holding still answers, still looks authoritative, and is wrong about the law. A dead link announces itself. Of the two, the collection of addresses is the more honest failure and the more frustrating one, and a reader who wants to know which kind of service they are using should look at whether the entries lead to documents on that host or somewhere else.

Coverage follows the funding, and the funding is on the page

One regional network is unusually candid about how it is paid for, and that candour is more useful than a coverage claim. The African platform describes itself as a programme of a university law department, jointly operated with a second organisation it names, and gives the European Union and two trusts as the funders of the platform. It states that the initial phase of the project ran from 2021 to 2024. It publishes African Union material itself and federates search across national collections held elsewhere, which is a precise account of what it does and does not hold.

It also illustrates how hard the count is to pin down. The body of that page puts the network at sixteen institutes. The footer of the same page lists twenty named entries under a heading for other African legal information institutions, one of which is the platform itself. The page does not reconcile the two figures and neither does this guide. What the discrepancy shows is that the boundary of a network like this is genuinely soft, and that a researcher asking whether a given country has an institute is asking a question the movement itself answers loosely.

Where the model is thin, the reason is usually the same and it is not indifference. Look at where this guide has cited things. Every settled field in the twenty country fact blocks here traces to a document published by the state whose law it describes, or to an archived capture of one, and not one of them traces to an institute. That is partly the method: this guide goes to the publisher. But it is also the pattern. In a state that publishes its own gazette well, an aggregator has little to add and no obvious way to fund adding it. In a state that publishes badly, an aggregator would be worth a great deal, and there is nothing to aggregate and nobody paying.

What that means before you rely on one

Three checks, and they take about as long as reading a home page. Find the page where the service says what it holds and to what date. Establish whether it publishes the documents or points at them. And find out who pays for it and whether the page says for how long, because a named funding period is more information about future coverage than most services give.

A fourth check applies only if you are building something rather than reading. Ten of the institute hosts tried while preparing this piece answered a plain automated request with a bot challenge rather than a page, both of the movement's global indexes among them. Nine of the ten said so with a refusal status. The tenth returned a success status carrying a page whose title said it was making sure the visitor was not a robot, which is the worse of the two answers, because a program reading only the status code records it as a page that loaded.

The material is genuinely free to a person with a browser and genuinely closed to a program, and nothing in the declaration addresses that distinction. It cannot be dismissed as somebody else's hosting decision either: the movement's own domain, asked for over a secure connection, answers with a certificate issued for one of those blocked institutes, and its declaration page could not be stored here from the live host at all.

The gap is worth sitting with, because free access was never only about price. The declaration commits its signatories to recognise the primary role of local initiatives in publishing their own national law. It is hard to read that beside a jurisdiction whose gazette section returned no records at all, or one whose issue numbering restarts and whose eleventh issue is missing from the list, and conclude that a challenge page was written with either of their readers in mind.

SOURCES

1. Declaration on Free Access to Law, Internet Archive capture of the movement's own page dated 2025-09-30. Free Access to Law Movement, captured by the Internet Archive. http://web.archive.org/web/20250930041303/http://www.fatlm.org/declaration/. Accessed 2026-08-31.
2. LII World Law by Source, the Cornell collection of world legal materials. Legal Information Institute, Cornell Law School. https://www.law.cornell.edu/world. Accessed 2026-08-31.
3. About AfricanLII, the platform's account of who runs it and what it publishes. AfricanLII, University of Cape Town, with Laws.Africa. https://africanlii.org/en/about/. Accessed 2026-08-31.
4. About Kenya Law, the organisational profile of the National Council for Law Reporting. National Council for Law Reporting, Kenya. https://kenyalaw.org/about/. Accessed 2026-08-31.
5. Bundesgesetz über das Bundesgesetzblatt 2004 (Bundesgesetzblattgesetz), consolidated text as at 30 August 2026. Bundeskanzleramt der Republik Österreich, Rechtsinformationssystem des Bundes. https://www.ris.bka.gv.at/GeltendeFassung.wxe?Abfrage=Bundesnormen&Gesetzesnummer=20002988&FassungVom=2026-08-30. Accessed 2026-08-30.

Reviewed 2026-08-31. 5 sources cited on this page.