1 of 8

Bridging Worlds: �Client and Law Firm Perspectives on AI and Innovation

Jonathan Hew

June 2026

Linklaters │ June 2026 │ 0

2 of 8

Some Questions

“The Changing Shape of Legal Work” – change or changes, shape or shapes …
Changing how, and why?
So what?

Linklaters │ June 2026 │ 1

One of the reasons “The Changing Shape of Legal Work” is a topic worth talking about because what’s really changing are lots of different, overlapping shapes, each representing different types of legal work done by different types of lawyers and adjacent professionals.

Conscious of the diverse audience today, I also hope it’s a topic that holds something for everyone.

For those who are new to legal practice, I hope it’ll be interesting to hear about the different shapes you might come across.

For those who have long committed themselves to one “shape”, I hope it’ll be interesting to hear about a few others.

And for those who have seen it all, I hope you’ll derive some validation from what I say, if not also pick up on something you can violently disagree with during the Q&A or afterwards.

3 of 8

Background

Disputes lawyer (Latham & Watkins, Hogan Lovells, Asylum Access)
In-house counsel and policymaker (Meta)
Innovation and AI lawyer (Linklaters)

Linklaters │ June 2026 │ 2

4 of 8

Law Firms

Depth – series of individual clients and matters.
Weak desire to systematise and reduce repeat work.
AI = breadth?

Linklaters │ June 2026 │ 3

Let’s start with the classic law firm piece of work.

The classic law firm piece of work can vary wildly in terms of size and subject matter, but compared to in-house work, it’s narrow and deep. It’s advising this client on this transaction, this case, or this specific issue. And then doing it again but for that other client.

This is for a reason – those lawyers are meant to be experts in their fields; that expertise comes at a (high) price; but you’re more likely to have your issue solved than without them. How do you know that’s the case? Trust, mostly.

The way law firms innovate, and do legaltech and tech generally, reflects this.

The drive to innovate comes in waves and silos – it’s primarily a response to external pressures from various clients, expressed in various ways and at various times, and is directed at increasing the efficiency of delivering work product for each of them while maintaining sufficient quality so as not to destroy trust.

For this reason, law firm tech stacks are often highly fragmented, innovations don’t scale, and top-down initiatives struggle to take root.

Generative AI, and especially the agentic layer built on it, is interesting because of its apparent flexibility – you may not need a bespoke solution, or as bespoke a solution, for each of Corporate, Finance, and Disputes as you did before.

Is it changing the shape of legal work though? “Kinda-sorta” – the clearest examples likely come from the transactional world and large-scale, AI-powered document review in the disputes world. That aside, it’s piecemeal – junior lawyers using AI to get at parts of processes, and senior lawyers using it for quick checks that they may previously have asked a junior to do. And that’s no surprise given how law firms work and are regulated.

What I’m most intrigued about is its potential to provide breadth – the ability to do things that you might not have been able to do otherwise, for example, to provide advice that accounts for a client’s historical positions, or the ability to provide more joined-up, cross-practice, truly “sectoral”, or even cross-sectoral, advice. I imagine that’ll create complications of its own, but it could meaningfully help bridge some of the divides that frustrate clients.

5 of 8

The Client View

Breadth – constant engagement with stakeholders and trade-offs between their various interests.
Strong desire to systematise and reduce repeat work.
AI = better breadth and some depth?

Linklaters │ June 2026 │ 4

The situation in-house is different.

By comparison, the classic in-house piece of work is wide and holistic, even in large and complex legal teams that have specialist coverage. It’s about advising a range of clients – stakeholders, really – not just on a given transaction, case, or issue, but how it interacts with others and what the right trade-offs are between the risks and opportunities that arise from each of them.

This too is for a reason – those lawyers are there not simply to be experts (which they often are), but to support the business’ wider interests. It’s called “General Counsel” for a reason.

The way they innovate, and do legaltech and tech generally, reflects this.

The drive to innovate is far more consistent and evenly-spread – unlike law firms, you want to reduce repeat work as much as possible so you can focus on the bigger picture. And you expect your law firms, often one of the biggest line items on the expenses sheet, to do the same (and they’ll oblige, in waves and silos that might be evening out a little more currently).

To be fair, I’m not sure this necessarily means that in-house legal tech stacks, innovation, and initiatives are better than law firms’ – I can imagine some being archaic are saddled with tech debt – but I feel the reasons are different, reflecting wider business structures rather than the specific business of the lawyers themselves.

Generative and agentic AI is interesting here too. In fact, I sometimes feel it’s more applicable in-house than it is to law firms, at least in the short run.

This is because something it seems to do particularly well is breadth. Thinking back to my time in-house, what I always wished I had more time for was synthesis and systematisation – whether of positions taken under a regulatory regime or across multiple regimes, or by developing frameworks from past experiences and applying them to the present and future.

It may also provide some of the depth that I would’ve had to go out to a law firm for. I doubt it’d replace them entirely, but it’d mean I needn’t rely on them as much and that I can start from a higher knowledge base and incur lower costs when I do.

6 of 8

The Overlap

Better means of law firm-client (and client-law firm) information sharing?
Better standardisation of client instructions, law firm work product, etc.?
Better structuring of data and measuring productivity?

Linklaters │ June 2026 │ 5

In addition, I feel that one way that the of legal work might change for both law firms and clients is how they overlap.

One example is how they share information with one another. It feels like AI can help clients convey to law firms the nature of the proverbial haystack that they’re going to have to find the needle in.

Another more prosaic example is the form of client instructions or law firm work product, which, frankly, is often all over the place.

A further issue touches both law firms and clients, though they’re likely to tackle it separately – their structuring of data and measurement of productivity. These are things that law firms and clients alike have either historically been bad at or are new to. Either way, there’s already significant pressure to get better at both, driven by a mixture of AI’s apparent capabilities and limitations, and supposed costs-savings and costs.

7 of 8

So What?

AI is promising for legal work, especially due to its flexibility.
That promise, and how it manifests, is context dependent.
Sensitivity to context is important for builders and users.

Linklaters │ June 2026 │ 6

8 of 8

Q&A

Linklaters │ June 2026 │ 7