Is it right that being able to afford a solicitor seems to improve the chances of avoiding prosecution?

[saismee]

It seems that OP revealed more mitigation to Manak in the thread that this splits from. People never seem to give us the full picture all in one go, just a string of excuses and half-truths to try and make themselves sound better, which is obviously unhelpful for themselves.

== Doublepost prevention - post automatically merged: 2 Jun 2026 ==

This thread seems to be quite relevant to this discussion!

Caught using younger brothers 11-15 Zip Card

I got caught with my younger brothers 11-15 zip card on the bus today. I lied to the ticket inspector that it was me and I lied about my age. Obviously he clocked on and then I ended up telling the truth about my age, name and that it was my brothers card. He confiscated it and asked for my...

www.railforums.co.uk

 

[35B]

It seems that OP revealed more mitigation to Manak in the thread that this splits from. People never seem to give us the full picture all in one go, just a string of excuses and half-truths to try and make themselves sound better, which is obviously unhelpful for themselves.

Which is where the premise of the original question is mistaken. A careful fact finding conversation, where someone is getting direct personal professional advice in real time, is very different from an online forum where the often essential Q&A is interrupted by a) the asynchronous nature of forum postings and b) the variety of voices in the discussion.

While money obviously helps here, the ability of a good professional advisor to elicit the key facts and then present them effectively is in itself likely to help get results.

 

[SECR 65]

being present in a CTA without a ticket

Sorry that this isn't quite on topic, but I've wondered for a while what a Compulsory ticketing area is?

 

[Haywain]

The thread linked to in post #61 is quite important to this discussion, as it shows that a warning can be obtained from TfL without a solicitor being involved.

 

[Watershed]

Sorry that this isn't quite on topic, but I've wondered for a while what a Compulsory ticketing area is?

An area defined as such under a Penalty Fares scheme, by means of signage advising that you are entering a CTA.

TfL declare all their stations and trains CTAs (though they fall under separate Byelaws where being present in a CTA without a ticket is punishable).

Chiltern and Nexus/T&W Metro are the only two other users of CTAs as far as I'm aware. Chiltern don't make all of their stations CTAs, just (some of?) the barriered ones. Presumably to avoid arguments as to "I was only going to the loo/trainspotting".

 

[SECR 65]

An area defined as such under a Penalty Fares scheme, by means of signage advising that you are entering a CTA.

TfL declare all their stations and trains CTAs (though they fall under separate Byelaws where being present in a CTA without a ticket is punishable).

Chiltern and Nexus/T&W Metro are the only two other users of CTAs as far as I'm aware. Chiltern don't make all of their stations CTAs, just (some of?) the barriered ones. Presumably to avoid arguments as to "I was only going to the loo/trainspotting".

Right, thank you. What is the significance of being in a CTA - does it mean certain things about requirements to have a ticket? I notice GTR NIP letters always talk about not having a valid ticket in "an area not designated as a compulsory ticketing area". Does this mean that, in a CTA, you can enter the station without a ticket provided you do not travel?

 

[blimmo]

Right, thank you. What is the significance of being in a CTA - does it mean certain things about requirements to have a ticket? I notice GTR NIP letters always talk about not having a valid ticket in "an area not designated as a compulsory ticketing area". Does this mean that, in a CTA, you can enter the station without a ticket provided you do not travel?

The other way around: you can be in stations that are not CTAs without a ticket but bylaw 18 makes it an offence to board a train without a valid ticket anywhere that is not a CTA.

 

[Watershed]

Right, thank you. What is the significance of being in a CTA - does it mean certain things about requirements to have a ticket? I notice GTR NIP letters always talk about not having a valid ticket in "an area not designated as a compulsory ticketing area". Does this mean that, in a CTA, you can enter the station without a ticket provided you do not travel?

If a station is designated a CTA, it means you can be issued with a Penalty Fare merely for being present in that area without a ticket. There is no requirement to prove that you travelled, or intended to travel. However, you cannot be prosecuted if it is in relation to National Rail services, since the existence of a CTA means that Byelaw 17 applies instead of Byelaw 18 - and there is no penalty for breaching Byelaw 17. For TfL stations/services, their Byelaws make a Byelaw 17 offence punishable.

If a station isn't designated a CTA, you can only be issued with a Penalty Fare if you have travelled or are on a train. You can only be prosecuted under Byelaw 18.

 

[SECR 65]

Thank you both for the clear explanation.

 

[island]

But with this message, a disclaimer is usually applied that it could potentially disqualify one from many professions.

This risk is considerably overstated. There is virtually no profession from which a single conviction for a minor fare evasion offence will disqualify someone, as long as the offence is proactively declared and accountability taken for it.

Not in this space, but AI used well can level this out.

Using AI to produce pleadings etc. used in legal proceedings risks being held in contempt of court, so I would strongly counsel against that.

Which is where the premise of the original question is mistaken. A careful fact finding conversation, where someone is getting direct personal professional advice in real time, is very different from an online forum where the often essential Q&A is interrupted by a) the asynchronous nature of forum postings and b) the variety of voices in the discussion.

and c) the bringing up of trivial or technical points of law often of little relevance to the matter under discussion

An area defined as such under a Penalty Fares scheme, by means of signage advising that you are entering a CTA.

TfL declare all their stations and trains CTAs (though they fall under separate Byelaws where being present in a CTA without a ticket is punishable).

This distinction is important. Under the National Rail Byelaws a CTA has to be designated under a Penalty Fares Scheme (which itself has no definition in law). Under TfL, a CTA in a station is designated by signage, in practice stickers on the ticket gates.

 

[ikcdab]

Using AI to produce pleadings etc. used in legal proceedings risks being held in contempt of court, so I would strongly counsel against that.

Why?

 

[island]

Why?

I would suggest you read the judgement of Sharp P in Ayinde v Haringey [2025] EWHC 1383.

 

[35B]

And the shredding of Pinsent Masons in Anthony Malcolm Cork & Anor v Mark Smith [2026] EWHC 1199 for a more recent example, which has an extended discussion of why a major law firm and/or one of its junior solicitors should or shouldn't be done for contempt.

Suffice it to say that the technology has a habit of inventing references, can lull users into a false sense of security, and requires significant cross-checking

 

[Fawkes Cat]

I would suggest you read the judgement of Sharp P in Ayinde v Haringey [2025] EWHC 1383.

A quick Google for this case brings up https://www.burges-salmon.com/articles/102krxw/professional-conduct-and-ai-ayinde-v-haringey/, which I haven’t fully read, but I think spells out the problem that AI is so keen* to please its user that it will make up an answer, with invented precedents and legislation, if it can't find true facts to do the job. That might be fine if you're an experienced and diligent lawyer - you'll spot the rubbish and get rid of it. But if you're not a lawyer, or not one of that sort, you may build your case on the rubbish - which at best will waste time for the court and opposing parties, and at worst lead to the case being decided wrongly.

It's not using AI to make your case that could be contempt of court: it's using that or any other tool without properly understanding what you are doing with it that could get you into trouble.

* Yes, I know. Anthropomorphic. But I hope it will do for the moment.

 

[35B]

A quick Google for this case brings up https://www.burges-salmon.com/articles/102krxw/professional-conduct-and-ai-ayinde-v-haringey/, which I haven’t fully read, but I think spells out the problem that AI is so keen* to please its user that it will make up an answer, with invented precedents and legislation, if it can't find true facts to do the job. That might be fine if you're an experienced and diligent lawyer - you'll spot the rubbish and get rid of it. But if you're not a lawyer, or not one of that sort, you may build your case on the rubbish - which at best will waste time for the court and opposing parties, and at worst lead to the case being decided wrongly.

It's not using AI to make your case that could be contempt of court: it's using that or any other tool without properly understanding what you are doing with it that could get you into trouble.

* Yes, I know. Anthropomorphic. But I hope it will do for the moment.

AI is the only tool currently in existence that could lead someone to make such a false submission. While the issue is with the human, the credence placed in AI (see especially Anthony Malcolm Cork & Anor v Mark Smith for how the mistaken reliance came about) is a feature of the tool that can lead even an intelligent and well trained professional astray

 

[pdsalford]

Can I just raise the suggestion that this question applies to the whole of the criminal justice system? It has always been that the better defence you can afford, gives you the best outcome. I'm not suggesting it's right for one moment, but it's how life works.

 

[island]

It's not using AI to make your case that could be contempt of court: it's using that or any other tool without properly understanding what you are doing with it that could get you into trouble.

Indeed, but in summary:

Using AI to produce pleadings etc. used in legal proceedings risks being held in contempt of court

 

[35B]

Can I just raise the suggestion that this question applies to the whole of the criminal justice system? It has always been that the better defence you can afford, gives you the best outcome. I'm not suggesting it's right for one moment, but it's how life works.

That said, one should keep an eye on the case of R vs Venables, where a tax specialist KC is being prosecuted for tax evasion.

In what may or may not be a related matter, his Oxfordshire home is on sale for £6.5m (see https://www.knightfrank.co.uk/prope...-studley-oxford-oxfordshire-ox33/cho012697348 for the estate agent's broochure)

 

[Dave W]

I think spells out the problem that AI is so keen* to please its user that it will make up an answer, with invented precedents and legislation, if it can't find true facts to do the job.

This is true of AI in many circumstances which make its application in the public sector quite tricky. But even here we can use it responsibly - I suspect in the legal world it's far too risky to make regular use (although, as the old acronym goes, IANAL)

 

[The exile]

Can I just raise the suggestion that this question applies to the whole of the criminal justice system? It has always been that the better defence you can afford, gives you the best outcome. I'm not suggesting it's right for one moment, but it's how life works.

And even if lawyers / solicitors etc were forced to charge a standard rate regardless of ability there would still be the issue of how long you could afford to engage them for - or, indeed, who had enough "inside knowledge" or "connections" to get access to the better ones. Unfortunately, absolute equality is unachievable - though that should not stop governments trying to minimise it.

 

[Egg Centric]

This is true of AI in many circumstances which make its application in the public sector quite tricky. But even here we can use it responsibly - I suspect in the legal world it's far too risky to make regular use (although, as the old acronym goes, IANAL)

As with any other tool you need a coming together of subject matter expertise and AI expertise. There are plenty of AI startups to do with the law, and certainly something like hallucinated cases is basically a solved problem in a correctly engineered solution. Getting the conclusion right etc though, not so much. N.B. that like Victorian bridge engineers, we're still figuring an enormous amount of stuff out, so there's going to be a lot of Tay Bridges (including what I'm currently working on imo but I didn't say that - but don't worry it's not going to cause a train crash or loss of life/limb!)

The lawyers though are just copying and pasting stuff into Copilot and being astonished it makes up stuff. That's like Victorian lawyers making the bridges out of sticks.

 

[Fawkes Cat]

The lawyers though are just copying and pasting stuff into Copilot and being astonished it makes up stuff

Also (and in line with where this tangent started) it's not just lawyers: it's easy enough for any of us to use AI to knock up something that looks lawyerly, but which on closer inspection is pants. So there's a risk from and to unrepresented defendants who are using AI to avoid the cost of a lawyer while still getting the supposed benefit.

 

[rmHawk765]

Unfortunately this is just the way the legal system works nowadays in most of the world, more money gets you a higher chance of getting what you want, it's simple as that. You always see big corporations involved because they can afford it.

 

[Puffing Devil]

As with any other tool you need a coming together of subject matter expertise and AI expertise. There are plenty of AI startups to do with the law, and certainly something like hallucinated cases is basically a solved problem in a correctly engineered solution. Getting the conclusion right etc though, not so much. N.B. that like Victorian bridge engineers, we're still figuring an enormous amount of stuff out, so there's going to be a lot of Tay Bridges (including what I'm currently working on imo but I didn't say that - but don't worry it's not going to cause a train crash or loss of life/limb!)

The lawyers though are just copying and pasting stuff into Copilot and being astonished it makes up stuff. That's like Victorian lawyers making the bridges out of sticks.

Framing the risk around the use of AI itself misses the mark. Courts are not penalising the use of technology; they are penalising the failure of verification. In Ayinde v Haringey Council, the issue was that a representative presented fictional authorities without checking them. The court's concern is the fundamental obligation not to mislead, a duty that remains identical whether a document was drafted by generative AI, copied from an unchecked textbook, or written by an unsupervised junior colleague.

The true risk lies entirely in submitting unchecked output within a formal legal document. When used responsibly as a starting point, AI is not a shortcut to contempt. The danger stems from overconfidence and blind trust in plausible text. Pope's phrase that a little learning is a dangerous thing fits perfectly here, as the real hazard arises when a user knows enough to generate a convincing brief but lacks the depth of knowledge required to spot a hallucinated citation.

 

[35B]

Framing the risk around the use of AI itself misses the mark. Courts are not penalising the use of technology; they are penalising the failure of verification. In Ayinde v Haringey Council, the issue was that a representative presented fictional authorities without checking them. The court's concern is the fundamental obligation not to mislead, a duty that remains identical whether a document was drafted by generative AI, copied from an unchecked textbook, or written by an unsupervised junior colleague.

The true risk lies entirely in submitting unchecked output within a formal legal document. When used responsibly as a starting point, AI is not a shortcut to contempt. The danger stems from overconfidence and blind trust in plausible text. Pope's phrase that a little learning is a dangerous thing fits perfectly here, as the real hazard arises when a user knows enough to generate a convincing brief but lacks the depth of knowledge required to spot a hallucinated citation.

There's a third layer, discussed in Anthony Malcolm Cork & Anor v Mark Smith, which is that the way in which AI works can undermine the objectivity of the person using it. That subliminal impact is I think the one that is quite specific to AI, because of how it is used and interacts

 

[Wolfie]

Why?

Not least because in a number of cases where it was done the AI software invented precedence.

Oh, and similar issues have arisen in other jurisdictions.

== Doublepost prevention - post automatically merged: 4 Jun 2026 ==

I would suggest you read the judgement of Sharp P in Ayinde v Haringey [2025] EWHC 1383.

Yup.

 

[Titfield]

Can I just raise the suggestion that this question applies to the whole of the criminal justice system? It has always been that the better defence you can afford, gives you the best outcome. I'm not suggesting it's right for one moment, but it's how life works.

This may be a question of semantics, but in my experience of sitting in an appeal tribunal, clients who are represented often do better because the paid representative is (a) able to put their case across in a far more coherent and logical way, highlighting positive aspects and at least trying to minimise negative aspects (b) actually knows the case inside and out (c) has prepared what to say and how to answer what are very often obvious questions a tribunal may ask (d) doesn't compound the issue with ill advised statements or even rants about the "injustice of the system".

On occasion unrepresented appellants seem unable to even muster the slightest of defences. At the least a good representative will know what to say that may strike a chord with the appeal panel and advance their case.

 

[Haywain]

clients who are represented often do better because the paid representative is

Detached from the outcome and therefore doesn't get emotionally distracted from presenting the case.

 

[Puffing Devil]

Detached from the outcome and therefore doesn't get emotionally distracted from presenting the case.

And it's an advocate's job to, err, advocate. Sure, you can go DIY, but it's a little different to repainting the lounge. And the long-term implications can be a lot more than the wrong shade and a few splashes on the carpet.

 

[ikcdab]

Not least because in a number of cases where it was done the AI software invented precedence.

Oh, and similar issues have arisen in other jurisdictions.

== Doublepost prevention - post automatically merged: 4 Jun 2026 ==


Yup.

So it's not using ai per se, that's ok. It more that you should make sure you are telling the truth and not inventing things. That sounds normal to me