Why we refuse Westlaw, Lexis and Practical Law

Those platforms sell licensed access to editorially maintained content: headnotes, annotations, practice notes, treatment signals, commentary written and updated by people who are paid to do it. The subscription your firm holds grants your fee-earners access on terms. It does not grant a software vendor the right to fetch, store, quote or redistribute that material through its own product.

A tool that scrapes those pages using a customer's credentials puts the customer in breach, not the vendor. That is an unattractive thing to build into a product sold to solicitors. It also tends to surface at the worst possible moment, which for a law firm is during a renewal conversation with the publisher.

The licence point is why we would not do it. The engineering point is why it would not work well anyway. Verification here is a comparison: the checker fetches the source, finds the cited passage, and confirms it appears word for word. Behind a login wall, the fetch returns a sign-in page. There is nothing to compare the quotation against.

A system that reports a pass without having read the source is precisely the failure this product exists to prevent. It would be worse than no check at all, because it would produce a green mark a supervising solicitor could reasonably rely on. An unverifiable authority presented as verified is the shape of the problem in Ayinde, not the answer to it.

The rule is simple. If the partner reading the memo cannot click the link and see the words, the tool does not cite it. Every one of the 74 approved sources is free and public, so verification and review use the same door. The associate who wrote it, the partner who signs it, the opponent who receives it and the judge who reads it all reach the same page.

There is a second effect firms notice a few weeks in. Because the sources are public, the client can check the memo too. That changes how an answer reads. It stops being an assertion backed by a subscription and becomes something the reader can test in about forty seconds, which is a different kind of document to put your name on.

No headnotes. No practice notes, no precedent bank, no editorially maintained commentary telling you how a line of authority developed. No publisher's citator signal on a case. For a litigator who checks whether a case is still good law with a commercial citator every morning, that is a real gap, and we are not going to argue otherwise on a marketing page.

So the honest recommendation is that a firm doing serious contentious work keeps its Westlaw or Lexis subscription. We are not a replacement for a research platform. We are a different tool doing something a research platform does not: producing a written answer where every authority carries a verdict, and refusing to produce one where the source cannot be opened.

The approved list is primary law and regulator material: legislation.gov.uk, BAILII, the National Archives Find Case Law service, EUR-Lex, the FCA Handbook, the ICO, the SRA, HMRC, gov.uk and others, covering England and Wales with EU, Irish and ECHR primary law alongside. That is enough for statutory work, for reported judgments, for regulatory questions and for most of what a small or mid-sized firm researches in a week.

It is not enough for everything, and the gaps are predictable. Unreported decisions that never reach a public source come back as not found even though they are real. Older material outside the public archives is out of reach. Anything turning on a publisher's editorial analysis is out of scope by definition, because that analysis is the licensed product.

Firms that get the most out of this run the two side by side and give each the work it is good at. Start with a research run: the question, the answer, the authorities table with verdicts, the panel showing what was searched. Read the not found and needs a check rows first, because that is where the risk sits and it is also where the reading is quickest.

Then take the four or five authorities the argument really rests on to your subscription platform and do what it is for. Read the headnote. Check the treatment. See who has cited it since. That is ten minutes on the cases that decide the advice rather than an hour on the ones that do not, and the licensed content is being used exactly as licensed: by a lawyer, in a browser, under your firm's own subscription.

Refusing paywalled sources is a constraint we chose, and constraints have costs that fall on the user rather than the vendor. You will occasionally get a not found on an authority you know perfectly well exists, because it lives behind a subscription or was never publicly reported. That is the system working as designed, and it is still an inconvenience on a Friday afternoon.

We also cannot tell you how any competitor handles this, and we will not guess. Harvey and Legora do not publish list pricing. CoCounsel is built on Thomson Reuters' Westlaw and Practical Law content, and Lexis+ AI on LexisNexis content, which means those products have licensed access to material we do not. If your practice depends on that corpus, that is a genuine reason to buy them rather than us.