Why junior lawyers shouldn’t be afraid of AI, by a law student

By on

We shouldn’t waste the abundance of opportunities out there


With tech becoming more and more advanced by the day, there’s a fear among law students that training contracts will be cut and law firm staff numbers will shrink. But penultimate London School of Economics law student Danial Alam is excited by new technologies and how they are changing the legal profession. Here’s why he thinks lawyers should be more optimistic about Artificial Intelligence (AI).

There’s a lot of buzz around AI at the moment and it is well placed.

In our homes, Amazon Alexa has become our virtual assistant. IBM Watson and Google Deepmind are pushing the boundaries on what computers can do and countless start-ups are finding exciting applications for AI that can be used today.

But with AI comes the question of automation. Companies like RAVN systems, Luminance and IBM Watson’s ROSS claim to offer document automation, smart contracting and more. A number of law firms — like Slaughter and May and Clifford Chance — have adopted these systems.

The fear is that if these technologies are going to cut time needed for due diligence then there will be a reduction in training contract numbers. So should aspiring lawyers be afraid?

In my opinion, no. Here is why I think that for at least the next ten years trainee solicitors are safe:

Change takes time

As with any technological shift, the introduction of AI into law firms will take time.

Take the development of the internet: what started out as connections between universities for very small amounts of data became the global network that we are so dependent on today. But, remember, that change took 30 years.

While the amount of investment and research in AI today is unprecedented, the society that we see in films like I, Robot and 2001: A Space Odyssey are still very distant futures. Our world is changing because of AI but such change in the legal industry is dependent on firms trusting the technology they are using, and that will only happen over time.

Work will become smarter

When we are at a stage where firms are willing to place their trust in AI systems to accurately perform actions previously done by some trainees, there will still be work for junior lawyers to do. But perhaps this is different work to what is currently being done.

Trainees will have more time to work on the complex challenges that firms face and will be able to develop skills in areas of work that AI will not be able to do, at least for the medium term. Strong written communication skills and the flexibility to adapt to new scenarios are examples.

More clients, and more work from existing ones

AI offers the opportunity for law firms to lower their cost for clients, as work will be able to be completed faster. But if this is true, then doesn’t it make sense for law firms to start lowering their intakes? Not exactly.

Firstly, as emerging markets become even more sophisticated, businesses will grow and those businesses will need lawyers. AI will allow firms to continue to service their existing clients while retaining the ability to take on new clients, as lawyers will be free to take on more work as processes become automated.

Also, if work can be done faster, law firms will be able to charge lower rates to their customers and this may not be a bad thing for firms. With lower rates more companies will be able to afford to pay for legal services and existing clients will be able to seek legal advice more readily. The revenue received from new work will counterbalance the reduction in price and will allow law firms to grow even more.

In ten years time, the legal industry will not be the same as we know it now, but that does not necessarily mean there will be less of a need for lawyers. In fact lawyers are needed now more than ever: experts will be called upon to navigate areas like self-driving cars, data protection in the Internet of Things and even AI itself. Aspiring lawyers should relish in the opportunities, rather than cower in fear.

For all the latest commercial awareness info, and advance notification of Legal Cheek’s careers events, sign up to the Legal Cheek Hub here.



No-one cares, Danial.



He’s in Denial.



I like photocopying


LLB student

Law students care! And should care. Change will happen with or without the non-carers.


Senior Associate, US firm London

Difficult to see how a law student’s insight into practice has much value here! There is no commentary here on what kind of work an AI might be able to do. I see no place for this in high value disputes. Its not necessarily a question of competence (although there is no substitute for human insight), its more an issue of trust. The first time an AI contributes to a major error, even on a ‘mundane’ task, the consequences will be catastrophic and this new tech will go back to the drawing board. Who will be accountable?



Noted. Thanks for your input. Get back to billing.


Senior associate at another US firm

Agreed with other US associate. It doesn’t affect you and me because we do complex stuff like negotiate non standard clauses which tech just can’t foreseeably ever do.

A note to young ‘uns – as fee pressure and tech drive corporate law further and further towards commodification, you’d be foolish not to seek out the complex stuff that will be the last to go over to the robots/DLA Sheffield….



because we do complex stuff like negotiate non standard clauses which tech just can’t foreseeably ever do.

I think this is just wrong. You may not be able to imagine it, but it’s definitely foreseeable – even if it may be 50 or 100 years away. You’re just not digesting enough science fiction.

‘HAL says this solution has a 65% probability of pleasing everyone’

Of course you are right on the other stuff – find the complex work, because the easier stuff will become more and more commodified.


Bloke who commented before

The word foreseeable means able to be forseen slash predicted. Humans might live until over 110 years old on average but it’s not ‘foreseeable’. It is foreseeable than low cost stuff will go because the tech exists and it is already happening.



I find it incredible that anyone thinks that negotiating clauses is beyond AI. As legal work goes it’s pretty ordinary stuff.

This will be done by AI apps within ten years.


It is foreseen/predicted. It doesn’t mater that it is foreseen/predicted in fiction, or in fact, because the technology is obviously moving in that direction and is close to becoming powerful enough to take the important next step. If a car can navigate itself in traffic, with all the numbskull (human) drivers (and cyclists/pedestrians) on the road, then how much more difficult, or how much of a technical leap do you think it is to make other sorts of rational decisions?

The only point you can make is to how far in the future that will be. That is something that we cannot actually predict. It may be like mining bases on the moon or colonies on Mars – hundreds of years in the future – or it could be like powerful personal communication (and tracking) devices in your pocket – here already.

Some of these sci-fi things don’t happen, because the basic economic rationale behind them doesn’t make any sense (like mining on the moon), but for others there are very good incentives, at least for the people who stand to benefit.


I clearly don’t think it is definitely going to happen. I’ve been in thousands of negotiations. Too many. I just don’t think the person above you knows what they’re talking about – but hey, it’s just an opinion and you know, they say opinions are like arseholes. Everyone has one. The reason why it won’t happen is because negotiation is a game of bluff, it’s tactics, it’s personalities. you have to have experience when to know when to press and when to stop. “AI” at the moment is really, really, really basic, it doesn’t even do thing which Excel is capable of doing to data within that programme – most of the software being provided for law firms are essentially Vlookup and some basic IF statements.


You, me and half the lawyers in the country have been in thousands of negotiations. Whoopeedoo.

If you don’t think AI can manage negotiations you’re both ignorant about AI and probably a shit negotiator.


Is chess or ‘go’ not about tactics? If a computer can learn those rules, then it can learn others.

Look at it this way: your ‘AI’ says ‘This is about the best deal I think we can get’. The other side’s ‘AI’ says ‘I concur’. Are you going to be arguing with them?

Junior Partner (non equity... yet)

I conclude from this discussion that you both have spare capacity. Please see me in my office. I have got a matter I would like both of you to action.



the people that made the code/algorithm behind the AI will be accountable.



How did Danial get interviewed by the BBC re the Brexit ruling at the Supreme Court?
A quick glance at his LinkedIn profile suggests that he doesn’t have a lot of “professional work experience”.



This post smacks of naivety.

Having worked in a legal department that focused on high-volume- low-cost work I can tell you the AI’s will take that first. So much can be automated that the cost will be paying for the servers to operate and to process client data. The last of the work in those departments will require minimal legal training.

Given that he is only looking at the next 10 years it is also safe to say that he is not looking at the long term. For a first-year law student, 10 would only allow them enough time to be 2 or so years post qualification. Not exactly a strong grounding in one’s career.



See my comment above regarding Danial’s non-existing professional experience.



This is what tech generally does, and law shouldn’t be any different:

Technology tends to bring down costs and make things quicker and easier. This tends to displace people wo are not willing/able to adapt to the tech. In turn this brings in new people who are able to adapt.

Some of these new people will be less skilled than the people displaced. These people will be paid less, but there may well be more of them. Some will be as skilled, and will have the tech skills. They will be paid more, and there will be relatively few of them.

There will be a degree of polarisation, between the cheap and tech savvy, and the expensive and tech savvy – the middle ranking will tend to fall away.

There will be space for some outside the tech-orientated sphere. But these people will need unique skills and be at the top of their game.



But, remember, that change took 30 years.

I think you are wrong on this, either because you don’t understand how quickly the internet was adopted and/or because you don’t understand that whateve AI is day zero is not today.

The first dial-up internet availability in the UK was in 1992. Within about five years it was becoming fairly mainstream, and the first mobile servies were available via WAP in 1999. By 2000 the first ADSL lines were being installed, and within a couple of years of that it too was mainstream.

That’s a significant change in 10 years – a lot less tha 30. Although everyone thinks Apple invented the smart phone in 2006, in fact there were precursors with email and web browsers.

With AI, the development probably started more than 50 years ago. ‘AI’ is just the latest buzzword for powerful computing and programming. If you think it’s something ‘new’, then you obviously have not read/seen ‘2001: a space odyssey’. Clearly a computer was perceived to be able to think and communicate – and the precursors to that story go back at least to 1948.

What you are seeing now is the fruit of all that work/imagination in the past – it’s likely that will start to have real ramifications over the next ten years, paticularly in jobs that require few intellectual skills (driving, cleaning, working in local government…). It isn’t going to take 30 years.


Latvka Da Blitz

Hey Doc. There’s only mentally I’ll kids here


Doc Litevsky.

Latvka. You are only a psychiatric nurse. Just notify me. But yes, there are only psychiatrically ill kids here. Thank you.


Comments are closed.