> ## Content Index
> Fetch the complete content index at: https://www.pradipkhakhar.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Why Judgement Alone Won’t Save Junior Lawyers From AI
- URL: https://www.pradipkhakhar.com/why-judgement-alone-wont-save-junior-lawyers-from-ai/
- Published: 2026-11-08T12:00:00.000Z
- Updated: 2026-11-08T11:59:59.000Z
- Author: Pradip Khakhar
- Tags: SQE Journey

Flex Legal recently asked me what single human skill junior lawyers need to master to thrive alongside AI. I said, it all comes down to client empathy and human judgement, and I meant it. Then in September, Professor Richard Susskind argued that judgement will not save lawyers from AI at all. My first reaction was to defend my answer. My second was to wonder if he had a point. This issue is about why judgement alone won’t save junior lawyers from AI, and what you need alongside it.

## 𝗪𝗵𝗮𝘁 𝗜 𝗦𝗮𝗶𝗱

This is the [answer](https://lnkd.in/p/ettZg8Bq?ref=pradipkhakhar.com) I gave Flex Legal for their Readiness Programme:

“It all comes down to client empathy. AI cannot tell you what matters most to the client. It can draft the letter, but the moment a client wants this resolved without burning the relationship, the legally strongest position becomes the wrong one. The number one skill is judgement, and the junior lawyers who thrive will not be the ones who use AI fastest, but the ones who know what it can never do.”

I still stand by most of it. Read it again, though, and look at what’s doing the work. My judgement only kicks in once the client has told me they want to keep the relationship. Somebody has to get that out of them first.

## 𝗪𝗵𝗲𝗿𝗲 𝗦𝘂𝘀𝘀𝗸𝗶𝗻𝗱 𝗣𝘂𝘀𝗵𝗲𝘀 𝗕𝗮𝗰𝗸

[Professor Susskind](https://www.legalfutures.co.uk/latest-news/susskind-judgement-will-not-save-lawyers-from-march-of-ai?ref=pradipkhakhar.com) takes on the objection lawyers reach for first. Judgement, he argues, is simply the tool humans use to manage uncertainty. Asking whether a machine can exercise judgement is “a little like asking whether submarines can swim”. The better question is whether machines can handle uncertainty, and he thinks they clearly can, drawing on far more data than the handful of matters any one lawyer has seen.

That’s uncomfortable reading for anyone who has built their pitch around judgement, and I include myself. I think he’s right about more than most lawyers want to admit. If judgement means weighing risk across known facts, a machine trained on millions of outcomes will get very good at it.

Notice the condition, though. **Known facts**.

## 𝗪𝗵𝗮𝘁 𝗨𝘁𝗮𝗵 𝗧𝗲𝗹𝗹𝘀 𝗨𝘀

Medicine is a step ahead of law here. Since late 2025, [Utah](https://www.beckershospitalreview.com/healthcare-information-technology/ai/utahs-ai-physician-experiment-expands-what-to-know/?ref=pradipkhakhar.com) has let an AI renew prescriptions for 192 common medications for chronic conditions as part of a state pilot. Last month it went further and approved another company’s AI to write new acne prescriptions without a doctor’s visit. In the early months of the renewal pilot, doctors reviewing the AI’s decisions agreed with it 91% of the time.

So the technology works. In May, the State told the company to strongly encourage patients to share their medical records before a refill. It stopped short of making that compulsory. The AI’s decision is only ever as good as what the patient chooses to hand over.

Not everyone was persuaded either. Utah’s own medical licensing board demanded the pilot be stopped on patient safety grounds. Whatever you make of that, a machine can be accurate and still not be trusted.

Lawyers face both problems. The client has to tell you the truth, and then the client has to act on what you advise.

## 𝗪𝗵𝗮𝘁 𝗪𝗮𝘀 𝗦𝗮𝗶𝗱 𝗔𝘁 𝗖𝗵𝗮𝗻𝗰𝗲𝗿𝘆 𝗟𝗮𝗻𝗲

At the [Junior Solicitor Summit](https://www.pradipkhakhar.com/junior-solicitor-summit-2026/) last month, the panel on the future lawyer kept coming back to ownership. Using AI responsibly came down to personal accountability, plus checking and verifying everything.

There was also a good exchange on learning. Taking the long view to learn something properly is often worth it, but firms rarely give juniors that time, so many end up experimenting on their own.

That second point stayed with me. If you’re teaching yourself the tools in the evenings, nobody is teaching you how to sit with a nervous client and get the full story. That was always learned by watching someone senior do it. As AI takes over the routine tasks that used to put juniors in the room, those chances to watch get rarer.

## 𝗧𝗿𝘂𝘀𝘁 𝗖𝗼𝗺𝗲𝘀 𝗕𝗲𝗳𝗼𝗿𝗲 𝗝𝘂𝗱𝗴𝗲𝗺𝗲𝗻𝘁

Clients rarely lie to their lawyers outright. They leave things out. The email they wish they’d never sent, or the deadline they’re facing with their own board. They hold it back because they’re embarrassed, or because they don’t yet believe you’re on their side.

Your judgement can only work on what’s in front of you. A brilliant analysis of half the picture rarely gives the best advice.

I saw this when I worked as a paralegal supporting a bankruptcy lawyer in the US. Our client was on a low income and came to us certain they needed to file, with plenty of paperwork to back it up. On paper it all made sense. It was only after we met in person and spent time getting to know each other that they shared something they’d been too embarrassed to mention. Filing was still the right outcome, but the way we could help them get there changed, and they left far happier than if we’d simply processed what they first asked for. Nothing about the legal analysis changed between those two meetings. What changed was how much they trusted us.

Trust like that gets built long before the meeting where you need it. It comes from small things, like returning the call when you said you would, or admitting what you don’t know yet. A client who feels understood tells you more.

## 𝗣𝗲𝗿𝘀𝘂𝗮𝘀𝗶𝗼𝗻 𝗖𝗼𝗺𝗲𝘀 𝗔𝗳𝘁𝗲𝗿

Then there’s the other end. You can be completely right and still lose the client. Good advice sitting unread in an inbox protects nobody.

Being correct is only part of why clients act on advice. They also need to understand it, and to believe the person giving it. That means explaining the risk in their terms rather than yours. It means knowing when they’re ready to hear that the strong position they want to take will cost them the relationship they care about.

This is where your interpersonal skills earn their keep, even when there’s no other side. 

One panellist at the summit put it neatly. Getting advice from AI is like a supermarket purchase, while buying from a person feels personal. The test is whether your client can see the value they got. A client who never acts on your advice never sees it.

## 𝗪𝗵𝘆 𝗜𝘁 𝗣𝗮𝘆𝘀

Empathy sounds soft until you see what it brings in. The client who trusts you comes back with the next matter and passes your name to a friend who needs a lawyer. When every firm has access to the same tools, being the person clients want to call is what sets you apart. I made the case for starting on this early in my issue on [starting business development before you qualify](https://www.pradipkhakhar.com/start-business-development-before-you-qualify-as-a-solicitor/).

## 𝗦𝗼, 𝗜𝘀 𝗝𝘂𝗱𝗴𝗲𝗺𝗲𝗻𝘁 𝗘𝗻𝗼𝘂𝗴𝗵?

No. I’d still put it at the centre, but it sits between two other skills. Before it comes getting the client to trust you with the whole story. After it comes getting them to act on what you advise. Susskind may well be right that machines will handle uncertainty better than we do. For now, they can’t get a frightened client to open up, and they can’t get a stubborn one to listen.

None of this is on the SQE. Practise it anyway. Every conversation with a manager, a customer or a colleague is a chance to notice when someone is holding back, and to ask the question that gets them talking.

## SQE Journey Will Be Back In Jan 2027

It is that time of year where I take my annual break from publishing. Today's edition is #93\. As we approach thanks giving in the US, I want to give a give shout out to all the professionals that have shared their knowledge and Insights here.

If you are studying and planning to take SQE 1 in January 2027, I wish you all the best!

SQE Journey will resume its weekly schedule starting Sunday, January 17, 2027\. We have some amazing guest interviews lined up.