Or it should be sealed for X years and then public record. Where X might be 1 in cases where you don't want to hurt an ongoing investigation, or 100 if it's someone's private affairs.

Nothing that goes through the courts should be sealed forever.

We should give up with the idea of databases which are 'open' to the public, but you have to pay to access, reproduction isn't allowed, records cost pounds per page, and bulk scraping is denied. That isn't open.

It's not easy to see who to believe. The MP introducing it claiming there is a "cover up" is just what MPs do. Of course it makes him look bad, a service he oversaw the introduction of is being withdrawn. The rebuttal by the company essentially denies everything. Simultaneously it's important to notice the government are working on a replacement system of their own.

I think this is a non-event. If you really want to read the court lists you already can, and without paying a company for the privilege. It sounds like HMCTS want to internalise this behaviour by providing a better centralised service themselves, and meanwhile all the fuss appears to be from a company operated by an ex-newspaper editor who just had its only income stream built around preferential access to court data cut off.

As for the openness of court data itself, wake me in another 800 years when present day norms have permeated the courts. Complaining about this aspect just shows a misunderstanding of the (arguably necessary) realities of the legal system.

https://www.tremark.co.uk/moj-orders-deletion-of-courtsdesk-...

They raise the interesting point that "publicly available" doesn't necessarily mean its free to store/process etc:

> One important distinction is that “publicly available” does not automatically mean “free to collect, combine, republish and retain indefinitely” in a searchable archive. Court lists and registers can include personal data, and compliance concerns often turn on how that information is processed at scale: who can access it, how long it is kept, whether it is shared onward, and what safeguards exist to reduce the risk of harm, especially in sensitive matters.

The counter claim by the government is that this isn't "the source of truth" being deleted but rather a subset presented more accessibly by a third party (CourtsDesk) which has allegedly breached privacy rules and the service agreement by passing sensitive info to an AI service.

Coverage of the "urgent question" in parliament on the subject here:

House of Commons, Courtsdesk Data Platform Urgent Question

https://www.bbc.co.uk/iplayer/episode/m002rg00

> ... the agreement restricts the supply of court data to news agencies and journalists only.

> However, a cursory review of the Courtsdesk website indicates that this same data is also being supplied to other third parties — including members of @InvestigatorsUK — who pay a fee for access.

> Those users could, in turn, deploy the information in live or prospective legal proceedings, something the agreement expressly prohibits.

Relatedly, there's an extremely good online archive of important cases in the past, but because they disallow crawlers in robots.txt: https://www.bailii.org/robots.txt not many people know about it. Personally I would prefer if all reporting on legal cases linked to the official transcript, but seemingly none of the parties involved finds it in their interest to make that work.

https://x.com/CPhilpOfficial/status/2021295301017923762

https://xcancel.com/CPhilpOfficial/status/202129530101792376...

"The government has cited a significant data protection breach as the reason for its decision - an issue it clearly has a duty to take seriously."

https://www.nuj.org.uk/resource/nuj-responds-to-order-for-th...

Then they start jailing people for posts.

Then they get rid of juries.

Then they get rid of public records.

What are they trying to hide?

https://www.bbc.co.uk/news/articles/c20dyzp4r42o

This principle applies directly to search engines and data processors. When Courtsdesk was deleted before reaching AI companies, the government recognized the constraint: AI systems can't apply this "completeness" standard. They preserve raw archives without contextual updates. They can't distinguish between "investigato" (under investigation) and "assolto" (aquitted). They can't enforce the court's requirement that information must be "correttamente aggiornata con la notizia dell'esito favorevole" (properly updated to reflect the favorable resolution of the proceedings).

The UK government prevented the structural violation the Cassazione just identified: permanent archival of incomplete truth. This isn't about suppressing information, it's about refusing to embed factually false contexts into systems designed for eternity.

you _really_ shouldn't be allowed to train on information without having a copyright license explicitly allowing it

"publicly available" isn't the same as "anyone can do whatever they want with it", just anyone can read it/use it for research

Quoting from an urgent question in the House of Lords a few days ago:

> HMCTS was working to expand and improve the service by creating a new data licence agreement with Courtsdesk and others to expand access to justice. It was in the course of making that arrangement with Courtsdesk that data protection issues came to light. What has arisen is that this private company has been sharing private, personal and legally sensitive information with a third-party AI company, including potentially the addresses and dates of birth of defendants and victims. That is a direct breach of our agreement with Courtsdesk, which the Conservatives negotiated.

https://hansard.parliament.uk/Commons/2026-02-10/debates/037...

In the United States our Constitution requires the government conduct all trials in public ( with the exception of some Family Court matters ). Secret trials are forbidden. This is a critical element to the operation of any democracy.

Ministry of Justice in UK has always struck me as very savvy, from my work in the UK civic tech scene. They're quite self-aware, and I assume this is more pro-social than it might seem.

> HMCTS acted to protect sensitive data after CourtsDesk sent information to a third-party AI company.

(statement from the UK Ministry of Justice on Twitter, CourtsDesk had ran the database)

but it's unclear how much this was an excuse to remove transparency and how much this actually is related to worry how AI could misuse this information

If the information is already accessibly to the public, even if what they claim about selling it to an AI company is true, the AI companies already has access to it if they want.

Shutting down the only working database is the proof point that perfect is the enemy of good.

Of course, this gives cover to the "well, we're working on something better" and "AI companies are evil"

Fine, shut it down AFTER that better thing is finally in prod (as if).

And if the data was already leaked because you didn't do your due diligence and didn't have a good contract, that's on you. It's out there now anyway.

And, really, what's the issue with AI finally having good citations? Are we really going to try to pretend that AI isn't now permanently embedded in the legal system and that lawyers won't use AI to write extremely formulaic filings?

This is either bureaucracy doing what it does, or an actual conspiracy to shut down external access to public court records. It doesn't actually matter which: what matters is that this needs to be overturned immediately.

A non-commercial initiative hosted by a group of Americans who wish to promote the cause of free information, but are themselves not interested in travelling to the UK, would be relatively immune to overreach by the British state. A US based service could take benefit of the First Amendment, the SPEECH Act, the fair use provision in US copyright law, etc - and simply ignore all the numerous British 'suppression orders', 'cessation notices', 'online safety' laws, etc.

If the Internet is indeed balkanizing - as sadly appears to be the case - let's squeeze some silver linings out of it.

I haven't confirmed it but I've seen journalist friends of mine complain also that the deletion period for much data is about 7 years so victims are genuinely shocked that they can't access their own transcripts even from 2018... Oh and if they can they're often extremely expensive.

This is why having decent standards in politics and opposition matters so much.

We all got together to vote out the last wankers, only to find that the current lot are of the same quality but in different ways.

And to think... the 'heads up their arses while trying to suppress a Hitler salute' brigade (Reform) are waiting in the wings to prove my point.