Need to know

• AI is creeping into courtrooms, with legal arguments backed by completely invented cases.

• Law firms are adapting, adding extra layers of checking and filtering AI generated work.

Welcome to TSL’s Weekly Edge, whether you’re aiming for a regional or high-street practice, or just want to get a feel for how law works in the real world beyond textbooks, you’re in the right place. 

No corporate jargon, no massive deals, just real useful information designed to give you that extra edge in your legal journey.

🧠Wilson’s Weekly Wisdom

Making mistakes does not disqualify you from success. If anything, it’s part of succeeding.

There’s this quiet pressure, especially in law, to look polished all the time. Get the grades. Say the right things. But the reality is much messier.

Throughout my training contract I’d receive feedback, receiving amendments with large sections crossed out. This felt like I was failing, however I soon learnt that people care far more about how you respond to a mistake than the mistake itself.

The people who quietly do well in this profession are not the ones who never get things wrong. They’re the ones who get better each time they do.

📣 Your Turn: Ask Us Anything (Almost)

Got a question that’s been quietly bugging you about the legal world, commercial awareness, training contracts, or how regional firms actually work day to day? Good. We want it.

Each month, we’ll pick a question and do a an editorial response. No fluff. No corporate waffle. Just honest, practical answers you can actually use in applications, interviews, and real conversations in firms.

If you’re wondering it, chances are someone else is too. So be brave, be curious, and send it in.

👉 Submit your question here!

💡Spotlight Article

A skeleton argument lands on a judge’s desk.

Looks sharp. Confident tone. Loads of case law neatly cited. It all screams “someone knows what they’re doing.”

Except for some of those cases? They’re not real.

Welcome to AI in legal proceedings.

🔎What’s happening? 

The scenario mentioned isn’t some imaginary whimsy. It’s actually happening.

Courts across England and Wales have started running into an uncomfortable reality: AI is increasingly used to draft legal arguments, inventing authorities that don’t exist and explaining legal principles in 30 seconds flat.

It isn’t checking Westlaw or reading judgments. It’s guessing what sounds convincing, based on patterns it’s seen before. Most of the time, it sounds confident enough that you’d never think to question it.

The issue is that AI doesn’t know the law, and that’s where the trouble starts.

Judges have now seen completely made-up cases, fake quotes from real judgments, and legal principles that just don’t exist being put before them.

Not once or twice, either, but enough times that senior judges have started issuing proper warnings.

Dame Victoria Sharp, President of the King’s Bench Division, described this as having “serious implications for the administration of justice and public confidence in the justice system,” and made clear that responsibility always sits with the lawyer, not the software.

In one case, it went far enough that a barrister was named in a public judgment after relying on AI‑generated authorities. The thinking was simple: if someone can’t spot something that basic, the public deserves to know.

So using AI isn’t the problem. Using it lazily is. 

Once it goes wrong, it’s not just embarrassing. It’s a lawyer’s name, reputation, and potentially career on the line.

Absolutely speaking, it’s a bad idea dressed up as efficiency!

❓ Why it matters to high street firms

It’d be easy to brush this off as just a “Bar problem” or something for advocates to worry about, but that’d be missing the point. High street firms are where it’s going to hit hardest, but also where these problems could be solved if handled properly.

Why?

Agree or not, many clients check AI before they ever pick up the phone.

The days of someone walking in with zero clue are gone. Now people ask AI to break down a tenancy dispute or divorce, turn up with letters they’ve already drafted, or run advice past LLMs just to sanity‑check it.

AI’s become a kind of sounding board, which isn’t entirely a bad thing. It can help people explain their situation better and feel confident enough to seek advice in the first place.

There’s a snag, however.

Sometimes, clients have the right idea. Other times, although written very confidently, those ideas are mostly or entirely wrong.

There’s also the bit perhaps no one wants to say out loud. AI can explain the basics quickly and for free. That chips away at the whole “premium for access to information” model.

So it’s rather unwise to rely on clients knowing nothing, because that’s old news.

But, and this is critical, explaining something isn’t the same as advising on it!

Context, judgment, risk, accountability. That’s still firmly a lawyer’s territory, so the role doesn’t disappear, it shifts.

Lawyers aren’t just the source of information anymore. They also filter and challenge, making sure everything holds up.

In practice, that means checking AI‑drafted letters and applications, stripping out anything hallucinated, fixing dodgy law, and stopping clients relying on something that sounds right but isn’t.

That’s not a step down. If anything, it’s a move deeper into risk management, exactly where the courts pay closer attention!

AI Hallucinations

Sounds dramatic, but it is fairly mundane, except when it presents a danger to the law.

An AI “hallucinates” when it confidently makes stuff up that looks right. The twist?

It doesn’t know it’s wrong, and it won’t give you a heads-up.